Index of sw-kernelinferencelibrary-public-generic-local/4f309dc0277a55fe4c4fbbf0c1df67299c94fa02/fmha/trtllm-gen
Name Last modified Size
../
include/ 16-Jul-2026 09:22 -
checksums.txt 16-Jul-2026 09:25 3.37 MB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 91.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 78.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 86.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 84.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 74.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 99.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 88.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 96.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 93.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 84.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 85.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 77.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 83.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 79.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 73.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 94.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 87.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 93.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 89.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 83.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 77.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 69.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 73.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 81.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 72.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 72.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 63.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 66.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 58.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 83.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 75.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 79.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 69.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 88.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 78.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 78.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 70.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 73.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 65.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 88.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 76.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 84.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 82.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 72.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 97.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 86.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 93.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 91.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 82.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 82.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 75.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 81.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 77.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 71.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 91.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 85.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 91.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 87.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 81.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 75.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 67.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 71.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 60.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 79.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 69.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 70.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 61.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 64.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 56.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 81.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 73.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 77.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 67.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 85.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 76.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 76.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 68.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 71.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 63.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 87.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 75.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 83.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 81.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 70.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 96.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 85.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 92.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 90.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 80.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 81.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 73.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 79.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 76.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 70.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 90.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 83.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 90.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 85.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 80.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 74.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 66.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 69.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 78.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 68.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 69.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 60.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 62.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 55.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 80.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 72.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 76.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 66.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 84.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 75.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 75.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 67.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 69.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 62.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 90.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 78.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 86.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 84.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 72.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 99.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 88.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 95.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 93.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 82.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 84.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 77.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 83.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 79.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 71.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 93.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 87.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 93.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 89.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 81.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 76.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 68.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 71.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 61.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 79.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 70.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 71.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 65.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 57.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 81.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 74.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 78.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 68.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 86.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 77.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 77.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 69.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 71.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 64.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 94.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 78.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 85.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 73.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 102.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 87.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 94.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 83.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 88.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 76.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 82.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 72.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 96.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 86.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 92.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 82.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 76.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 68.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 63.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 76.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 71.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 65.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 60.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 82.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 81.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 75.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 73.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 83.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 81.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 72.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 70.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 92.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 75.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 83.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 71.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 85.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 92.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 81.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 86.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 74.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 80.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 70.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 94.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 84.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 90.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 80.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 73.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 69.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 66.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 61.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 74.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 69.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 63.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 58.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 79.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 78.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 73.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 71.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 80.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 78.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 69.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 90.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 74.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 81.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 70.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 99.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 84.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 91.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 80.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 84.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 73.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 79.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 93.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 83.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 79.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 72.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 68.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 65.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 60.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 73.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 67.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 62.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 57.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 78.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 77.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 71.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 70.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 79.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 77.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 68.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 67.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 93.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 77.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 85.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 73.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 102.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 87.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 94.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 83.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 87.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 76.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 82.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 70.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 96.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 86.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 92.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 80.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 74.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 70.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 67.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 62.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 75.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 69.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 63.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 59.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 80.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 79.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 73.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 81.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 79.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 70.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 77.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 73.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 87.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 83.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 76.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 72.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 86.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 82.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 67.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 63.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 64.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 60.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 74.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 73.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 71.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 70.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 75.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 71.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 85.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 80.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 74.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 70.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 84.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 80.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 65.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 61.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 62.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 58.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 72.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 71.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 69.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 68.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 74.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 84.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 79.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 72.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 69.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 82.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 78.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 64.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 60.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 61.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 57.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 71.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 70.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 68.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 67.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 77.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 73.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 87.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 82.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 76.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 71.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 86.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 80.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 66.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 62.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 63.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 59.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 73.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 70.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 97.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 82.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 80.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 90.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 89.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 77.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 76.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 106.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 92.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 90.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 100.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 98.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 87.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 85.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 92.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 80.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 79.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 87.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 84.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 76.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 75.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 90.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 97.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 94.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 86.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 85.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 84.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 76.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 76.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 72.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 65.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 62.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 86.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 76.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 77.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 68.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 71.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 68.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 61.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 59.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 90.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 82.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 83.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 79.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 72.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 92.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 83.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 83.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 75.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 77.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 75.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 68.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 66.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 95.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 80.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 78.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 88.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 87.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 75.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 73.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 104.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 90.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 88.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 97.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 96.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 85.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 83.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 89.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 78.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 77.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 85.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 82.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 74.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 73.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 98.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 88.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 87.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 95.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 92.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 84.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 83.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 82.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 74.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 74.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 70.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 63.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 60.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 83.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 74.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 75.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 66.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 69.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 66.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 59.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 57.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 88.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 80.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 81.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 77.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 70.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 67.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 90.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 81.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 81.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 73.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 75.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 73.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 66.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 64.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 94.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 78.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 77.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 87.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 86.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 74.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 72.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 103.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 89.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 87.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 96.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 95.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 84.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 82.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 88.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 77.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 75.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 84.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 81.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 73.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 72.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 97.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 87.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 85.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 94.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 90.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 83.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 81.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 81.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 73.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 73.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 69.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 58.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 82.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 72.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 74.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 65.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 67.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 65.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 58.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 55.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 86.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 79.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 79.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 75.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 69.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 65.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 88.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 79.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 80.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 74.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 71.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 65.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 62.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 98.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 82.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 81.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 91.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 90.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 75.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 76.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 106.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 92.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 91.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 100.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 99.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 85.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 86.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 92.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 81.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 79.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 88.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 85.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 75.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 73.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 91.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 98.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 94.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 85.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 83.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 83.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 75.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 75.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 69.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 65.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 61.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 84.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 75.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 77.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 67.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 68.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 66.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 60.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 58.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 89.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 82.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 82.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 76.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 91.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 82.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 83.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 74.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 75.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 73.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 67.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 65.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 114.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 91.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 99.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 86.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 104.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 96.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 112.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 89.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 97.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 84.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 102.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 94.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 111.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 88.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 96.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 82.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 93.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 114.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 92.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 99.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 84.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 103.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 95.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 97.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 107.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 105.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 94.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 118.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 126.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 126.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 114.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 102.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 99.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 93.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 124.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 116.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 123.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 120.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 113.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 92.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 80.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 102.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 91.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 93.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 83.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 84.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 76.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 111.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 99.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 107.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 92.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 117.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 102.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 107.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 94.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 99.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 87.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 92.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 101.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 100.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 88.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 112.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 120.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 120.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 109.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 90.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 94.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 87.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 118.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 110.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 117.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 114.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 107.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 86.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 75.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 96.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 85.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 87.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 78.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 80.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 71.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 105.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 94.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 100.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 87.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 110.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 97.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 101.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 89.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 93.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 82.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 89.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 98.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 97.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 86.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 110.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 118.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 117.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 106.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 87.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 94.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 91.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 85.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 116.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 108.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 115.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 111.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 105.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 83.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 72.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 93.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 83.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 84.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 75.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 78.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 68.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 102.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 92.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 98.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 85.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 108.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 95.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 98.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 87.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 90.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 80.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 91.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 99.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 99.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 88.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 112.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 120.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 119.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 108.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 89.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 93.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 86.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 118.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 110.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 117.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 113.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 107.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 84.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 74.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 94.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 84.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 86.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 76.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 77.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 70.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 103.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 93.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 99.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 86.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 109.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 97.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 100.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 89.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 92.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 82.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 97.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 93.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 117.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 114.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 92.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 115.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 112.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 89.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 81.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 82.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 79.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 103.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 102.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 97.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 99.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 91.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 88.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 111.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 108.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 86.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 110.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 107.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 82.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 76.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 77.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 73.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 97.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 97.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 91.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 94.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 88.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 85.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 109.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 105.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 87.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 84.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 107.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 104.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 80.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 74.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 74.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 71.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 94.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 94.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 88.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 91.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 91.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 87.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 111.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 107.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 89.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 86.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 109.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 106.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 81.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 75.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 76.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 72.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 96.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 95.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 90.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 92.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 101.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 99.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 111.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 111.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 97.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 96.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 122.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 120.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 131.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 131.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 118.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 116.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 100.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 98.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 107.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 105.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 95.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 131.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 120.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 118.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 128.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 125.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 117.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 115.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 92.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 84.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 80.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 107.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 96.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 98.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 88.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 91.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 87.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 79.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 77.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 118.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 105.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 110.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 106.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 95.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 91.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 121.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 107.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 112.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 99.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 104.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 102.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 90.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 88.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 96.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 94.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 105.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 105.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 92.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 90.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 116.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 114.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 125.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 125.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 112.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 111.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 94.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 92.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 101.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 99.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 91.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 125.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 114.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 112.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 122.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 119.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 111.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 109.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 89.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 85.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 78.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 74.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 100.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 90.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 92.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 82.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 84.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 81.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 73.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 71.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 111.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 101.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 104.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 100.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 90.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 86.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 115.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 102.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 106.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 94.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 98.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 95.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 85.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 83.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 93.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 91.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 102.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 103.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 89.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 88.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 114.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 112.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 122.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 123.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 110.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 108.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 91.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 99.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 97.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 88.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 87.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 123.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 112.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 110.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 119.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 117.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 108.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 107.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 87.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 82.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 76.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 72.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 98.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 87.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 89.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 80.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 81.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 78.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 71.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 109.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 98.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 101.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 97.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 88.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 84.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 112.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 100.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 103.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 92.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 96.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 93.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 83.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 81.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 94.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 104.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 105.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 92.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 90.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 116.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 114.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 125.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 112.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 111.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 94.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 92.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 101.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 99.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 91.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 89.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 125.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 114.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 112.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 122.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 119.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 111.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 109.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 89.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 85.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 78.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 74.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 100.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 89.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 91.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 82.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 81.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 79.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 73.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 71.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 111.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 101.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 103.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 100.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 90.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 87.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 114.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 102.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 105.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 94.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 95.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 93.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 86.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 83.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 182.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 178.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 179.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 176.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 148.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 165.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 143.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 162.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 169.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 165.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 166.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 163.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 134.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 153.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 129.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 150.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 163.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 159.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 161.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 157.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 130.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 148.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 125.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 144.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 185.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 181.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 183.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 179.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 151.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 141.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 146.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 137.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 172.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 168.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 170.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 166.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 137.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 128.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 133.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 124.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 167.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 162.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 164.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 161.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 133.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 123.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 129.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 119.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 80.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 68.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 75.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 74.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 65.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 86.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 73.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 80.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 79.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 71.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 75.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 67.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 73.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 70.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 65.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 81.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 72.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 78.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 75.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 70.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 71.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 62.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 63.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 55.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 74.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 62.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 63.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 55.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 60.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 52.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 75.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 67.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 67.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 60.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 77.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 66.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 67.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 59.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 64.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 56.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 78.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 65.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 72.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 72.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 63.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 84.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 71.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 78.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 77.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 68.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 73.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 65.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 71.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 68.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 63.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 79.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 70.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 76.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 73.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 68.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 68.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 60.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 61.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 53.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 71.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 60.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 61.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 52.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 58.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 50.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 73.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 65.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 65.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 57.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 75.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 64.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 65.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 57.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 62.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 54.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 76.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 64.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 71.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 71.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 62.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 82.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 70.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 76.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 76.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 67.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 63.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 70.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 66.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 61.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 78.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 69.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 75.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 72.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 67.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 67.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 59.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 60.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 52.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 70.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 59.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 60.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 51.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 57.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 49.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 72.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 64.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 64.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 56.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 74.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 63.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 64.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 55.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 61.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 53.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 79.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 66.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 74.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 74.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 64.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 85.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 72.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 79.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 78.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 70.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 75.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 65.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 73.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 67.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 63.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 80.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 72.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 78.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 74.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 69.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 69.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 61.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 62.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 54.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 72.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 61.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 62.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 53.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 59.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 51.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 73.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 66.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 66.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 76.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 65.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 65.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 58.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 63.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 55.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 83.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 67.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 73.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 65.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 89.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 72.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 79.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 70.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 78.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 66.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 72.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 64.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 84.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 71.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 77.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 69.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 62.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 54.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 69.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 59.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 59.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 52.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 74.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 68.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 66.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 59.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 73.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 65.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 63.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 57.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 80.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 65.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 71.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 63.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 86.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 70.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 76.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 68.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 76.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 64.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 70.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 62.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 82.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 69.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 75.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 67.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 67.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 60.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 59.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 52.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 66.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 57.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 57.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 50.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 71.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 66.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 63.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 57.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 70.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 62.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 61.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 55.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 79.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 63.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 70.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 61.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 85.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 69.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 75.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 67.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 74.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 62.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 61.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 81.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 68.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 74.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 66.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 65.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 58.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 51.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 65.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 56.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 56.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 49.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 70.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 65.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 62.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 56.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 69.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 61.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 60.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 54.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 82.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 67.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 73.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 65.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 88.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 71.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 78.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 69.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 77.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 64.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 72.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 62.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 83.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 70.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 76.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 67.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 60.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 60.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 53.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 67.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 58.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 58.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 51.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 72.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 66.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 64.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 58.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 71.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 63.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 62.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 56.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 66.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 64.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 72.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 70.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 65.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 64.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 71.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 60.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 54.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 58.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 52.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 65.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 59.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 62.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 57.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 64.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 62.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 67.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 63.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 61.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 69.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 67.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 58.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 52.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 56.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 50.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 62.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 57.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 60.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 55.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 63.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 61.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 66.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 60.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 67.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 65.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 57.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 51.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 55.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 49.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 61.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 56.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 59.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 54.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 66.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 64.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 71.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 64.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 62.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 70.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 68.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 59.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 53.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 57.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 51.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 63.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 58.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 61.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 56.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 87.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 71.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 69.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 79.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 79.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 69.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 67.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 93.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 77.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 75.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 84.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 84.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 74.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 73.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 82.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 70.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 68.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 77.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 75.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 68.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 67.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 88.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 76.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 74.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 83.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 80.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 73.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 72.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 77.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 69.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 69.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 65.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 55.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 78.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 66.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 70.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 59.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 65.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 62.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 55.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 53.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 82.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 74.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 73.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 69.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 63.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 59.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 82.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 70.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 74.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 63.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 69.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 66.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 59.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 57.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 84.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 69.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 67.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 77.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 77.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 67.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 65.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 90.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 74.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 72.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 82.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 82.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 72.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 70.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 80.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 68.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 66.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 75.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 73.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 66.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 64.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 86.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 74.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 80.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 78.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 71.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 70.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 75.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 67.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 67.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 63.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 56.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 53.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 76.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 64.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 68.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 57.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 63.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 60.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 53.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 51.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 80.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 72.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 71.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 67.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 61.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 57.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 79.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 68.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 72.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 61.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 67.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 64.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 57.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 55.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 83.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 68.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 66.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 75.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 76.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 65.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 64.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 89.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 73.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 71.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 81.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 81.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 71.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 69.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 78.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 67.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 65.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 74.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 72.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 65.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 63.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 84.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 72.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 70.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 79.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 77.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 70.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 74.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 65.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 65.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 62.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 55.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 52.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 74.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 63.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 67.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 56.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 62.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 59.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 52.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 49.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 78.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 70.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 70.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 66.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 60.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 56.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 78.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 67.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 71.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 60.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 65.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 63.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 56.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 54.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 87.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 69.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 70.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 79.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 79.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 67.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 68.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 92.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 76.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 75.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 84.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 84.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 74.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 73.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 82.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 69.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 67.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 78.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 72.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 66.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 65.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 88.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 76.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 74.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 82.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 80.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 73.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 72.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 76.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 68.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 66.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 63.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 58.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 54.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 77.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 65.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 67.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 62.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 60.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 54.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 52.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 81.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 73.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 70.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 67.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 62.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 59.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 81.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 70.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 71.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 63.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 66.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 64.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 59.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 57.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 102.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 76.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 84.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 73.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 96.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 88.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 74.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 82.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 71.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 93.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 86.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 98.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 73.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 81.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 92.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 84.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 101.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 76.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 84.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 73.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 95.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 87.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 93.01 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 86.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 90.76 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 83.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 88.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 81.86 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 89.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 82.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 92.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 86.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 90.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 84.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 87.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 81.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 88.76 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 82.53 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 101.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 95.28 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 99.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 92.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 97.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 90.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 99.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 92.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 105.01 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 93.39 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 95.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 85.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 91.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 80.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 93.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 81.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 104.78 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 100.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 96.53 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 91.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 92.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 87.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 93.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 87.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 113.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 100.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 105.42 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 91.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 101.75 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 87.01 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 103.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 88.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 87.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 81.61 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 83.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 76.03 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 81.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 73.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 82.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 74.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 86.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 81.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 82.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 76.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 80.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 74.29 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 81.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 75.16 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 95.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 90.04 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 92.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 85.29 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 90.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 83.17 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 92.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 85.04 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 95.76 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 88.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 93.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 86.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 91.26 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 84.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 92.26 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 85.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 95.31 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 73.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 80.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 68.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 97.17 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 88.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 71.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 77.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 89.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 77.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 69.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 66.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 58.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 73.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 65.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 64.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 56.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 79.09 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 71.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 93.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 71.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 78.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 66.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 95.17 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 86.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 69.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 75.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 65.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 87.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 74.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 67.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 64.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 56.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 71.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 63.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 61.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 54.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 76.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 69.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 91.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 69.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 76.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 65.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 93.42 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 84.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 73.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 64.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 85.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 71.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 66.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 61.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 54.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 68.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 61.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 59.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 52.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 73.67 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 67.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 93.81 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 71.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 78.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 67.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 95.42 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 86.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 70.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 75.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 66.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 88.00 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 73.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 67.61 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 63.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 56.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 70.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 63.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 61.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 54.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 75.42 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 69.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 72.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 68.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 71.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 67.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 65.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 58.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 95.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 88.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 62.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 56.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 70.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 66.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 69.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 65.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 63.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 56.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 92.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 86.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 60.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 54.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 64.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 67.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 64.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 60.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 54.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 90.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 84.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 58.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 52.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 71.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 66.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 69.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 66.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 63.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 56.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 91.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 85.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 60.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 54.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 75.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 71.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 74.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 70.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 70.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 60.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 104.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 97.78 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 67.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 57.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 73.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 72.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 68.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 58.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 102.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 95.53 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 64.97 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 55.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 72.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 67.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 70.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 66.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 65.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 56.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 100.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 93.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 62.47 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 53.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 74.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 70.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 73.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 69.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 66.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 58.46 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 101.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 95.03 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 63.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 56.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 107.33 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 76.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 84.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 71.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 101.58 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 92.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 105.46 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 74.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 82.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 99.33 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 90.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 103.71 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 72.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 80.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 67.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 96.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 88.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 106.33 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 75.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 83.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 70.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 98.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 90.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 107.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 95.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 98.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 87.47 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 94.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 82.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 96.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 83.71 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 98.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 94.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 96.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 93.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 107.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 103.00 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 89.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 82.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 87.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 80.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 90.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 86.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 88.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 85.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 99.28 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 93.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 80.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 75.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 78.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 72.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 86.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 82.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 84.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 81.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 95.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 89.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 76.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 70.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 74.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 88.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 84.39 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 86.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 83.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 96.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 90.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 78.40 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 71.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 75.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 101.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 98.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 99.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 96.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 116.67 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 102.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 93.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 84.46 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 89.97 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 82.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 93.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 89.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 91.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 88.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 108.29 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 93.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 86.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 75.58 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 83.22 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 73.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 89.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 85.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 87.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 84.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 104.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 89.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 82.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 71.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 79.22 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 68.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 91.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 87.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 89.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 86.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 106.75 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 90.97 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 81.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 72.83 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 78.97 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 70.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 87.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 81.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 83.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 76.03 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 81.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 73.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 82.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 74.78 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 89.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 71.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 78.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 92.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 84.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 70.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 76.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 68.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 86.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 73.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 66.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 65.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 59.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 70.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 63.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 63.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 56.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 75.06 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 68.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 85.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 67.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 73.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 65.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 87.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 80.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 66.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 72.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 64.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 82.39 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 70.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 62.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 62.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 54.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 67.04 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 59.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 59.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 52.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 71.81 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 64.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 83.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 65.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 72.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 63.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 86.31 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 78.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 64.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 70.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 62.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 80.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 67.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 60.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 59.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 53.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 64.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 58.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 57.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 51.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 69.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 63.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 85.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 67.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 74.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 65.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 88.31 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 80.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 66.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 72.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 65.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 82.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 69.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 62.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 61.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 54.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 66.54 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 59.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 59.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 52.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 71.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 64.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 71.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 69.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 70.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 68.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 64.40 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 59.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 86.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 80.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 62.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 57.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 66.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 64.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 65.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 63.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 61.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 54.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 82.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 75.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 58.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 52.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 64.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 63.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 64.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 62.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 59.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 53.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 80.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 73.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 56.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 51.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 67.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 65.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 66.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 64.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 61.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 54.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 81.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 74.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 58.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 52.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 74.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 72.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 73.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 71.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 69.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 59.46 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 95.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 89.67 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 66.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 57.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 69.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 67.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 68.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 66.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 65.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 55.71 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 92.06 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 84.92 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 62.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 53.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 67.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 65.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 67.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 65.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 63.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 53.71 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 89.81 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 82.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 60.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 51.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 70.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 68.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 69.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 68.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 64.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 56.21 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 91.81 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 84.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 61.97 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 54.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 103.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 75.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 83.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 72.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 98.28 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 89.58 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 98.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 70.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 79.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 67.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 94.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 85.83 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 97.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 69.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 77.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 66.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 91.75 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 83.83 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 72.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 80.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 69.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 94.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 86.33 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:28 111.67 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:25 98.95 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:25 144.32 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:26 131.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:27 114.38 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:27 101.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:25 113.38 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:27 100.73 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 145.30 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:25 132.70 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:24 116.22 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:24 103.55 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:27 111.80 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:23 99.07 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:25 144.45 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:23 131.77 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 114.50 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:26 101.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 113.51 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:23 100.85 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:26 145.43 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:24 132.82 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 116.34 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:26 103.55 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 113.52 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:24 94.21 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:25 145.84 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:23 127.12 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:24 116.37 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:25 96.91 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:28 115.23 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:24 95.88 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:26 147.65 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:23 128.05 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:24 118.07 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:25 98.69 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:24 113.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:24 94.21 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:24 145.96 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:24 127.24 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:25 116.49 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:28 97.03 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:24 115.35 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:27 96.00 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:26 147.65 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:25 128.17 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 118.20 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:24 98.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 140.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 114.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 123.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 109.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 129.64 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 113.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 120.63 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 108.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 117.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 111.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 104.31 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 97.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 113.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 107.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 100.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 94.29 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 138.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 111.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 120.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 106.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 126.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 110.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 117.82 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 105.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 114.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 108.71 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 101.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 110.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 104.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 97.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 91.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 136.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 110.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 119.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 105.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 125.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 108.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 116.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 104.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 112.85 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 107.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 99.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 93.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 109.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 102.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 96.13 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 90.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 139.85 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 113.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 122.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 106.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 128.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 112.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 119.93 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 105.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 115.08 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 109.44 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 102.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 111.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 105.06 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 98.13 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 92.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 114.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 108.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 112.68 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 107.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 102.86 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 97.53 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 99.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 94.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 111.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 105.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 109.87 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 105.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 99.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 94.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 96.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 91.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 109.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 104.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 108.33 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 103.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 98.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 93.31 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 94.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 90.01 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 113.94 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 106.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 112.47 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 105.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 101.15 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 95.55 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 97.39 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 92.12 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 118.31 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 112.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 116.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 111.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 108.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 92.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 103.78 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 115.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 109.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 113.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 108.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 105.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 100.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 86.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 113.96 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 108.37 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 112.36 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 107.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 104.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 88.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 99.29 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 84.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 118.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 110.24 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 116.74 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 109.29 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 105.13 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 91.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 100.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 87.93 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 148.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 119.01 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 128.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 112.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 140.00 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 126.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 145.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 116.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 125.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 110.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 137.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 123.96 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 144.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 114.78 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 124.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 108.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 135.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 122.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 148.21 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 119.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 128.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 110.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 138.55 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 125.47 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 149.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 145.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 147.14 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 143.74 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 136.17 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 131.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 128.22 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 127.72 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 142.61 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 138.22 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 140.33 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 136.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 128.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 124.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 121.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 120.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 139.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 135.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 137.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 133.61 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 124.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 121.92 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 117.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 117.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 142.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 137.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 139.65 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 136.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 126.82 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 123.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 119.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 119.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 153.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 148.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 150.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 147.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 139.24 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 119.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 133.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 114.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 146.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 141.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 143.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 140.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 131.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 112.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 126.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 108.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 143.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 138.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 140.92 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 137.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 128.24 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 110.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 122.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 105.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 145.94 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 141.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 143.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 139.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 130.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 112.79 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 123.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 108.55 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 225.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 219.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 222.21 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 217.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 189.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 205.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 180.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 200.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 209.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 203.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 206.08 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 201.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 172.87 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 190.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 164.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 185.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 202.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 196.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 198.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 194.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 167.71 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 183.21 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 159.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 178.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 230.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 223.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 226.53 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 221.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 192.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 176.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 185.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 171.85 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 214.01 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 207.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 210.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 205.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 175.71 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 161.22 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 169.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 155.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 206.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 200.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 203.00 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 198.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 170.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 155.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 164.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 150.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 127.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 100.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 108.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 98.00 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 119.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 99.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 106.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 96.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 111.17 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 101.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 96.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 85.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 104.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 91.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 93.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 82.76 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 124.86 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 97.63 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 105.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.19 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 116.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 96.64 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 103.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 94.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 108.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 98.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 93.35 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 82.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 101.64 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 89.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 90.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 80.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 123.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 96.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 104.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 93.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 114.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 95.11 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 102.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 92.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 106.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 96.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 91.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 81.08 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 100.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 87.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 88.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 78.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 126.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 99.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 106.92 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 96.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 117.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 97.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 105.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 95.33 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 109.19 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 98.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 94.12 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 83.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 102.53 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 89.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 91.19 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 80.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 99.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 97.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 98.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 96.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 94.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 85.24 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 91.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 82.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 96.79 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 94.47 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 95.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 93.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 91.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 82.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 88.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 80.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 95.26 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 92.94 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 94.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 91.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 90.26 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 81.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 87.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 78.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 98.82 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 96.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 97.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 93.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 83.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 90.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 80.78 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 103.63 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 101.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 102.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 100.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 100.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 84.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 96.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 81.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 100.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 98.37 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 99.85 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 97.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 97.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 81.36 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 93.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 78.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 99.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 96.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 98.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 95.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 79.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 92.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 77.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 103.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 100.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 102.12 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 99.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 97.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 83.12 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 94.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 80.29 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 137.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 104.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 114.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 101.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 134.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 119.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 135.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 102.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 111.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 98.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 131.08 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.13 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 133.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 100.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 110.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 97.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 129.65 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 115.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 137.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 104.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 114.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 100.87 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 132.79 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 118.41 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 105.19 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 94.00 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 109.62 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 98.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 105.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 93.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 110.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 99.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 104.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 93.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 109.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 99.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 115.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 104.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 120.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 110.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 138.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 123.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 133.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 125.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 116.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 148.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 127.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 138.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 128.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 120.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 129.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 121.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 130.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 115.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 115.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 136.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 125.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 135.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 118.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 118.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 128.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 112.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 124.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 108.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 127.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 114.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 136.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 121.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 115.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 100.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 115.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 101.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 135.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 121.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 127.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 112.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 132.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 120.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 140.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 103.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 118.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 105.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 132.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 115.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 120.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 108.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 143.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 120.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 131.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 123.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 113.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 124.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 114.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 121.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 110.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 107.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 131.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 118.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 127.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 113.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 111.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 122.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 106.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 116.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 98.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 121.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 109.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 127.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 113.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 109.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 95.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 106.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 92.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 129.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 116.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 120.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 126.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 115.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 132.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 118.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 113.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 99.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 110.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 129.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 120.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 117.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 140.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 114.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 120.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 107.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 121.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 108.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 117.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 107.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 102.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 128.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 113.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 122.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 110.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 106.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 119.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 103.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 111.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 94.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 118.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 106.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 122.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 107.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 105.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 92.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 101.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 87.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 126.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 113.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 115.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 98.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 123.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 111.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 127.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 113.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 108.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 95.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 104.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 91.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 131.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 112.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 122.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 119.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 105.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 141.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 117.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 128.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 122.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 109.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 123.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 110.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 119.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 109.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 104.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 130.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 115.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 125.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 112.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 108.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 120.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 105.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 111.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 95.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 118.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 106.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 123.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 109.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 107.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 93.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 101.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 89.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 127.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 114.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 116.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 100.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 123.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 112.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 128.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 114.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 110.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 97.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 106.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 93.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 141.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 121.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 130.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 114.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 148.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 132.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 120.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 128.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 114.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 137.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 128.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 112.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 120.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 105.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 130.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 114.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 113.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 100.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 131.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 119.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 136.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 114.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 122.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 107.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 144.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 127.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 113.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 120.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 106.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 132.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 121.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 106.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 112.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 97.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 121.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 106.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 104.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 92.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 126.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 114.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 132.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 109.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 117.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 102.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 140.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 123.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 107.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 116.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 101.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 128.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 118.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 103.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 107.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 92.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 115.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 101.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 98.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 87.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 122.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 110.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 134.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 110.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 119.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 104.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 142.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 125.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 109.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 118.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 102.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 130.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 119.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 104.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 108.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 93.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 116.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 102.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 99.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 88.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 123.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 112.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 126.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 115.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 131.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 120.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 120.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 109.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 125.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 114.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 117.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 106.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 122.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 111.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 117.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 106.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 122.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 112.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 139.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 102.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 107.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 90.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 128.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 93.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 100.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 98.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 89.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:23 128.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 116.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 93.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 80.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 100.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 87.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 96.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 85.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 90.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 77.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 145.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 127.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 138.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 132.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 120.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 153.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 132.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 143.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 134.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 124.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 136.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 126.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 134.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 121.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 120.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 141.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 130.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 139.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 124.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 123.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 136.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 120.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 128.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 110.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 138.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 126.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 140.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 125.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 120.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 106.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 118.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 104.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 140.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 126.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 132.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 116.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 143.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 131.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 144.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 130.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 123.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 109.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 122.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 109.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 140.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 120.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 130.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 127.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 112.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 148.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 124.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 135.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 129.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 131.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 118.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 126.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 117.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 112.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 136.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 123.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 132.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 119.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 116.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 130.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 114.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 120.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 102.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 132.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 121.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 131.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 118.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 115.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 101.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 109.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 96.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 134.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 121.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 125.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 106.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 137.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 126.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 137.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 123.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 118.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 104.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 114.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 99.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 136.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 115.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 125.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 123.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 107.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 144.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 119.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 130.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 126.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 111.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 128.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 113.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 122.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 113.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 106.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 133.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 117.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 127.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 116.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 110.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 126.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 111.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 114.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 98.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 129.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 117.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 126.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 112.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 111.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 97.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 104.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 91.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 131.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 117.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 120.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 102.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 134.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 122.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 131.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 117.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 114.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 100.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 109.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 95.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 139.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 118.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 128.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 125.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 110.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 147.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 122.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 134.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 128.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 115.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 131.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 116.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 125.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 115.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 110.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 135.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 121.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 131.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 114.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 128.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 113.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 117.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 100.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 130.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 119.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 128.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 115.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 113.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 106.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 93.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 133.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 120.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 122.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 107.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 135.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 124.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 134.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 120.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 116.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 102.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 114.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 99.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 149.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 126.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 136.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 119.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 148.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 132.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 131.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 144.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 118.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 128.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 111.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 142.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 126.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 125.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 141.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 113.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 123.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 106.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 139.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 123.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 122.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 143.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 116.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 127.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 109.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 141.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 125.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 123.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 105.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 93.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 110.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 99.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 104.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 93.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 109.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 99.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 101.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 86.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 104.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 89.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 100.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 89.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 104.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 92.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 111.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 97.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 114.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 100.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 146.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 151.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 163.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 137.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 144.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 152.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 152.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 165.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 139.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 145.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 133.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 148.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 158.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 126.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 142.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 138.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 149.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 160.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 128.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 143.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 141.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 132.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 122.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 115.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 145.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 130.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 145.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 127.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 109.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 135.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 122.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 144.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 135.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 124.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 122.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 149.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 134.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 163.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 149.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 125.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 112.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 138.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 124.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 131.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 136.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 147.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 122.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 129.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 137.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 139.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 151.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 124.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 132.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 133.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 142.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 111.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 127.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 124.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 136.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 147.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 113.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 130.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 122.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 116.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 106.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 100.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 133.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 117.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 146.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 132.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 110.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 93.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 121.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 107.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 125.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 122.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 109.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 107.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 137.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 119.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 151.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 135.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 111.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 96.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 124.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 110.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 124.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 128.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 140.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 115.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 122.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 131.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 132.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 145.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 118.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 111.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 126.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 135.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 104.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 120.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 118.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 130.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 141.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 107.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 124.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 114.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 109.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 100.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 94.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 126.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 109.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 139.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 102.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 87.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 114.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 100.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 117.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 116.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 103.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 101.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 130.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 114.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 144.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 130.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 105.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 91.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 118.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 104.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 125.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 129.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 141.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 115.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 122.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 131.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 133.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 146.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 119.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 126.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 111.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 126.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 136.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 105.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 121.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 118.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 131.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 141.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 108.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 115.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 109.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 101.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 95.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 125.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 108.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 138.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 123.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 102.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 88.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 113.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 100.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 118.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 115.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 104.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 102.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 130.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 113.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 144.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 128.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 105.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 91.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 118.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 105.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 142.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 126.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 145.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 130.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 123.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 111.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 126.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 114.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 115.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 103.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 118.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 107.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 115.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 103.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 118.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 107.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 164.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 118.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 127.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 135.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 114.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 152.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 115.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 124.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 124.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 113.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 152.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 141.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 116.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 103.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 125.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 111.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 124.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 110.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 111.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 100.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 154.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 154.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 168.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 143.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 147.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 156.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 156.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 169.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 144.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 148.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 141.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 152.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 163.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 132.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 146.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 143.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 154.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 165.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 133.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 147.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 153.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 140.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 133.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 123.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 149.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 134.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 166.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 149.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 132.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 115.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 140.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 155.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 139.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 136.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 127.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 153.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 138.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 169.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 153.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 130.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 117.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 143.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 127.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 139.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 140.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 152.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 127.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 134.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 141.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 144.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 156.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 129.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 136.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 126.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 138.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 147.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 116.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 132.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 128.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 141.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 151.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 135.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 134.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 124.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 117.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 108.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 137.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 118.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 152.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 135.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 115.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 99.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 128.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 111.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:24 137.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 126.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 120.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 111.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 141.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 122.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 156.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 139.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 117.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 101.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 131.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 113.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 132.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 133.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 145.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 120.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 126.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 135.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 137.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 150.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 123.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 130.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 119.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 131.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 140.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 109.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 124.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 122.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 135.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 145.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 112.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 128.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 126.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 116.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 111.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 102.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 129.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 112.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 143.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 128.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 107.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 93.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 120.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 104.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 129.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 120.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 114.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 106.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 134.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 116.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 149.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 133.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 96.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 124.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 108.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 133.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 134.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 146.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 121.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 126.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 136.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 138.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 151.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 124.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 131.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 120.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 131.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 141.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 110.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 125.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 123.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 136.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 147.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 113.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 129.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 127.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 113.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 103.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 130.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 114.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 144.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 128.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 108.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 94.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 118.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 106.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 129.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 120.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 116.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 107.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 134.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 119.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 149.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 133.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 111.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 97.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 122.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 110.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 100.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 86.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 104.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 89.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 101.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 89.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 104.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 92.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 160.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 167.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 155.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 154.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 158.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 164.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 153.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 131.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 120.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 139.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 127.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 136.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 124.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 127.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 118.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 109.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 97.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 111.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 100.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 108.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 100.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 111.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 103.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 153.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 154.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 154.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 152.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 159.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 150.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 150.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 145.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 153.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 146.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 162.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 153.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 153.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 144.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 151.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 140.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 129.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 125.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 134.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 128.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 136.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 131.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 123.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 118.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 130.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 117.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:28 139.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 135.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 126.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 126.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 132.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 118.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 116.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 112.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 123.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 115.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 123.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 118.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 112.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 106.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 117.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 104.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 126.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 124.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:26 115.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 113.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 120.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 106.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 184.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 193.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 179.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 170.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 182.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 182.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 190.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 178.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 158.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 147.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 168.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 155.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 172.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 165.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:23 149.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 154.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 154.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 145.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 158.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 149.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:27 161.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:24 152.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 138.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 128.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 141.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:27 131.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:25 124.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:28 115.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:26 127.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:25 118.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 161.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 160.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 159.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 158.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 159.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 152.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 158.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 151.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 156.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 148.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 156.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 144.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 136.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 131.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 138.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 133.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 139.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 125.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 135.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 122.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 140.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 131.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 137.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 123.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 123.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 127.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 120.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 125.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 114.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 122.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 109.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 129.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 125.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 112.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 231.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 222.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 228.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 219.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 227.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 240.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 220.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 224.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 239.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 216.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 202.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 192.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 224.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 213.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 190.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 184.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 198.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 194.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 224.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 216.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 186.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 183.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 193.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 184.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 198.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 189.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 188.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 204.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 181.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 194.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 209.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 187.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 176.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 167.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 201.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 188.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 164.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 155.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 179.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 167.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 204.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 189.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 166.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 158.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 178.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 169.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 183.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 174.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 173.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 188.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 166.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 179.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 194.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 171.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 162.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 150.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 185.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 171.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 149.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 139.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 169.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 155.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 192.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 178.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 155.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 144.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 210.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 203.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 211.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 316.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 327.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 311.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 181.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 171.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 196.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 182.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 190.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 183.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 171.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 165.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 236.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 227.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 232.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 223.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 231.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 245.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 224.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 229.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 244.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 221.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 206.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 196.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 229.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 221.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 195.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 188.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 202.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 198.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 226.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 220.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 191.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 186.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 197.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 188.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 203.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 193.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 193.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 208.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 185.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 199.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 213.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 191.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 181.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 171.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 205.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 192.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 170.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 159.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 183.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 171.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 208.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 194.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 172.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 182.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 173.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 187.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 178.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 178.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 193.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 170.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 183.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 199.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 175.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 166.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 154.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 189.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 175.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 155.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 142.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 173.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 160.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 196.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 182.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 148.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 100.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 88.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 106.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 94.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 100.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 88.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 106.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 94.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 111.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 99.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 105.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 122.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 108.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 118.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 111.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 104.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 132.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 111.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 122.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 114.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 107.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 113.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 107.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 116.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 103.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 103.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 122.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 110.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 120.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 106.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 106.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 113.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 99.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 107.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 95.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 104.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 119.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 107.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 101.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 86.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 102.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 89.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 120.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 109.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 111.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 99.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 123.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 111.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 125.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 112.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 105.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 90.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 105.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 93.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 119.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 113.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 108.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 99.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 129.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 108.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 118.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 111.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 103.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 111.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 102.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 111.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 100.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 98.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 120.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 107.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 117.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 104.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 97.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 102.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 90.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 114.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 102.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 115.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 103.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 99.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 84.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 97.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 85.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 118.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 106.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 107.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 94.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 120.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 108.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 121.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 108.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 103.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 88.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 101.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 89.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 101.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 111.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 107.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 97.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 128.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 105.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 116.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 110.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 101.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 109.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 100.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 109.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 99.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 96.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 118.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 104.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 115.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 102.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 100.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 108.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 95.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 100.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 88.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 112.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 100.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 113.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 101.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 97.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 82.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 95.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 83.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 116.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 105.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 104.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 92.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 118.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 107.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 119.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 106.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 101.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 86.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 99.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 87.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 121.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 104.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 114.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 110.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 100.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 131.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 109.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 120.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 113.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 105.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 112.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 112.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 103.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 99.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 121.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 108.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 118.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 106.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 104.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 110.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 97.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 102.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 90.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 114.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 101.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 115.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 85.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 97.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 86.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 106.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 107.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 95.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 120.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 108.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 121.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 109.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 103.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 88.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 101.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 90.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 125.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 106.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 115.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 102.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 133.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 116.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 104.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 114.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 101.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 123.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 111.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 99.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 104.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 92.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 114.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 100.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 100.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 88.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 106.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 122.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 102.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 112.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 98.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 130.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 114.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 101.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 110.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 97.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 120.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 108.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 96.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 99.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 87.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 110.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 95.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 95.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 83.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 114.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 103.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 120.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 100.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 109.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 96.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 128.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 112.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 99.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 108.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 95.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 119.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 107.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 94.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 97.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 85.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 108.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 93.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 93.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 81.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 112.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 101.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 124.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 103.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 112.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 99.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 131.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 116.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 101.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 111.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 98.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 122.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 109.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 96.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 99.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 87.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 110.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 96.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 95.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 83.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 115.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 103.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 114.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 104.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 122.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 111.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 112.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 102.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 119.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 108.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 111.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 100.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 117.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 106.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 112.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 101.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 119.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 108.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 238.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 84.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 92.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 96.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 83.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 230.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 83.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 90.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 89.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 81.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 229.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 216.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 85.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 72.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 91.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 78.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 86.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 75.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 82.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 69.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 129.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 112.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 123.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 108.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 137.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 116.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 127.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 120.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 111.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 121.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 111.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 121.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 110.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 107.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 127.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 115.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 125.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 112.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 110.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 120.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 106.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 110.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 98.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 128.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 116.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 124.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 112.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 107.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 92.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 105.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 93.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 125.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 113.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 114.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 102.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 134.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 122.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 129.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 110.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 96.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 110.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 97.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 126.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 108.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 119.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 115.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 104.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 134.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 112.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 124.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 118.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 108.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 118.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 107.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 117.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 107.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 103.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 124.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 111.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 122.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 107.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 104.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 106.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 94.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 125.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 113.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 120.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 108.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 104.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 89.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 101.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 89.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 123.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 111.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 110.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 98.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 131.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 119.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 125.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 113.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 108.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 93.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 105.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 93.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 125.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 106.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 116.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 113.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 102.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 133.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 110.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 121.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 116.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 105.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 116.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 105.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 115.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 105.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 101.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 123.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 109.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 120.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 109.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 105.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 116.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 102.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 104.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 92.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 124.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 112.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 118.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 106.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 103.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 88.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 98.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 87.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 121.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 109.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 108.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 130.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 118.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 123.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 111.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 106.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 91.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 103.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 90.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 129.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 110.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 121.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 117.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 106.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 136.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 115.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 126.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 119.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 110.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 120.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 109.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 119.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 109.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 105.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 126.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 114.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 124.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 112.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 109.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 119.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 105.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 107.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 95.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 126.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 114.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 122.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 109.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 106.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 94.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 102.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 90.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 124.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 112.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 112.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 102.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 132.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 120.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 127.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 114.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 109.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 97.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 108.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 96.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 133.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 111.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 123.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 106.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 133.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 121.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 120.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 131.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 107.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 119.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 102.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 130.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 119.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 118.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 129.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 105.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 116.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 100.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 128.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 116.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 133.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 109.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 120.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 104.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 133.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 120.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 118.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 104.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 89.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 110.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 94.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 105.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 88.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 110.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 94.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 115.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 100.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 121.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 105.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 126.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 110.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 132.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 116.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 120.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 105.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 126.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 111.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 117.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 101.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 123.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 107.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 126.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 110.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 132.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 116.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 120.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 104.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 126.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 110.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 117.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 101.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 123.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 107.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 137.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 122.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 143.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 127.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 132.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 117.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 137.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 122.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 128.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 113.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 134.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 118.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 107.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:27 96.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 113.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 102.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 107.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 96.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 113.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 101.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 111.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 98.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 116.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 103.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 110.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 99.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 115.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 104.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 160.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 161.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 174.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 147.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 155.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 160.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 163.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 174.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 146.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 156.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 147.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 159.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 136.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 153.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 147.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 161.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 135.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 155.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 145.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 134.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 159.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 147.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 171.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 159.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 137.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 124.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 151.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 141.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 141.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 134.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 158.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 146.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 170.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 158.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 132.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 123.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 148.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 140.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 141.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 145.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 155.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 128.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 138.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 143.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 147.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 159.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 129.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 141.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 128.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 142.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 152.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 117.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 137.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 130.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 145.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 155.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 119.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 139.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 126.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 115.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 143.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 127.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 154.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 139.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:24 116.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 104.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 132.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 121.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 128.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 117.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 145.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 130.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 157.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 143.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 117.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 105.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 135.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 124.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 132.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 135.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 146.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 119.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 129.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 135.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 140.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 151.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 122.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 133.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 119.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 133.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 143.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 108.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 127.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 122.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 137.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 147.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 111.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 131.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 116.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 105.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 133.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 118.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 144.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 130.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 107.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 95.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 123.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 113.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 120.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 109.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 137.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 122.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 149.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 135.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:24 109.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 98.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 127.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 116.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 171.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 166.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 178.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 155.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 159.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 171.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 166.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 177.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 154.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 159.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 158.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 163.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 175.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 144.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 158.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 158.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 164.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 174.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 143.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 158.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 155.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 145.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 162.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 148.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 174.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 160.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 144.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 130.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 154.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 142.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 152.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 147.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 160.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 148.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 173.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 161.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 139.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 133.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 151.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 142.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 152.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 148.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 159.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 136.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 142.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 154.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 151.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 162.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 137.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 144.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 139.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 145.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 155.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 125.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 140.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 141.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 149.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 158.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 126.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 142.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 136.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 124.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 145.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 128.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 157.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 140.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 123.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 110.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 135.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 122.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:24 138.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 126.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 148.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 131.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 160.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 144.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 124.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 112.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 137.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 125.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 143.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 138.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 150.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 127.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 132.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 146.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 143.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 154.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 129.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 136.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 130.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 136.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 146.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 116.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 131.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 133.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 140.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 151.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 119.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 134.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 127.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 115.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 136.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 119.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 148.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 131.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 114.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 101.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 126.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 113.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 130.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 118.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 140.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 123.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 152.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 136.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 117.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 104.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 130.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 117.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 147.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 144.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 133.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 119.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 124.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 112.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 108.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 110.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 99.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 152.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 185.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 163.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 197.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 145.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 178.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 154.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 185.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 166.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 198.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 147.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 177.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 146.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 146.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 151.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 184.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 160.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 194.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 144.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 177.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 153.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 184.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 163.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 195.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 146.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 177.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 151.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 138.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 173.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 170.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 163.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 148.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 187.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 181.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 142.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 139.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 146.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 133.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 167.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 164.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 152.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 140.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 172.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 164.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 151.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 188.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 182.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:24 142.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 139.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 146.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 134.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 166.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 164.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 140.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 158.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 151.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 169.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 133.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 150.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 144.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 161.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 155.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 173.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 136.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 153.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 118.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 119.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 139.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 157.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 148.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 167.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 133.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 150.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 142.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 160.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 152.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 136.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 152.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 139.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 125.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 154.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 144.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 151.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 135.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 167.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 155.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 122.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 112.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 134.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 120.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 148.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 137.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 142.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 129.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 154.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 146.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 154.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 140.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 169.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 158.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 121.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 114.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 136.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 123.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 148.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 140.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 135.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 146.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 146.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 158.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 128.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 139.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 139.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 149.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 151.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 162.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 132.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 141.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 106.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 108.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 134.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 145.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 143.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 155.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 128.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 138.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 138.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 148.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 148.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 159.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 131.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 141.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 134.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 120.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 142.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 132.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 145.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 131.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 154.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 142.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 108.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 100.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 129.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 115.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 136.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 125.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 137.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 124.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 145.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 135.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 150.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 135.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 158.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 146.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 110.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 102.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 132.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 119.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 139.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 128.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 137.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 145.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 133.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 133.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 135.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 142.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 131.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 136.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 123.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 144.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 130.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 139.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 127.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 131.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 121.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 155.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 188.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 167.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 200.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 148.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 181.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 158.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 188.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 169.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 201.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 150.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 180.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 154.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 154.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 154.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 187.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 163.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 197.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 147.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 180.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 156.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 187.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 166.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 198.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 149.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 180.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 153.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 139.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 178.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 160.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 166.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 149.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 190.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 149.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 138.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 148.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 133.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 172.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 154.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 154.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 140.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 175.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 161.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 167.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 151.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 191.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 172.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 149.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 138.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 149.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 134.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 169.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 154.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 143.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 161.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 155.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 173.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 136.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 154.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 147.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 164.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 159.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 177.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 139.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 156.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 125.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 127.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 142.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 160.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 151.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 170.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 135.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 153.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 145.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 163.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 156.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 174.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 139.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 155.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 141.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 126.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 157.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 139.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 153.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 136.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 170.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 149.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 129.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 117.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 136.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 120.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 151.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 132.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 144.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 129.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 157.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 140.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 157.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 139.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 172.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 152.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 129.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 118.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 139.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 123.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 151.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 133.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 138.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 149.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 149.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 161.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 131.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 142.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 142.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 153.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 154.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 165.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 135.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 144.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 114.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 116.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 137.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 148.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 146.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 158.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 130.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 141.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 141.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 151.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 151.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 162.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 134.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 144.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 136.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 121.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 144.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 125.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 148.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 131.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 157.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 136.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 116.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:23 104.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 131.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 115.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 139.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 119.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 140.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 124.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 147.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 129.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 152.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 135.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 161.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 140.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 118.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 107.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 134.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 118.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 141.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 122.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 166.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 163.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 151.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 133.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 138.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 125.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 118.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 122.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 109.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 166.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 206.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 185.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 225.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 159.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 200.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 167.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 207.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 188.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 226.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 161.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 200.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 149.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 164.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:25 153.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 164.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 163.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 204.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 178.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 220.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 158.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 198.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 165.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 205.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 181.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 221.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 159.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 198.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 164.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 150.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 194.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 191.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 180.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 169.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 211.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 208.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 159.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 162.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 143.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 157.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 157.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 144.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 187.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 184.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 164.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 152.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 190.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 191.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 181.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 171.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 208.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 208.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:25 164.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 158.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:27 147.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 158.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 158.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 146.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 183.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 184.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 151.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 175.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 170.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 195.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 145.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 168.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 155.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 178.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 175.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 199.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 148.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 132.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 131.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 137.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 134.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 149.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 172.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 163.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 188.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 143.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 166.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 153.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 176.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 168.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 192.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 147.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 170.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 149.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 135.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 169.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 159.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 164.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 154.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 186.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 176.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 142.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 137.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 127.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 126.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 142.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 129.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 163.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 152.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 153.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 139.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 170.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 163.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 169.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 158.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 188.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 180.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 146.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:23 138.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 130.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 128.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 146.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 133.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 163.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 156.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 145.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 159.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 165.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 180.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 139.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 153.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 149.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 163.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 170.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 184.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 143.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 156.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:23 125.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 117.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 129.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 119.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 143.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 157.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 158.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 173.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 137.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 151.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 147.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 161.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 163.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 178.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 142.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 155.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 143.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 130.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 155.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 144.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 158.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 148.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 171.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 164.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 134.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 121.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:27 119.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 111.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 137.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 123.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 148.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 138.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 147.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 134.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 159.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 148.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 163.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 153.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 176.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 168.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 138.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 124.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 123.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 114.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 141.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 128.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 151.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 141.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 154.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 163.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 150.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:27 145.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 152.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 152.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 160.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 149.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 150.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 139.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 162.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 147.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 169.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 160.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 139.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 146.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 146.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 137.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 169.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 210.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 188.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 229.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 162.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 203.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 170.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 210.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 191.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 230.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 164.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 203.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:25 156.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 172.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 160.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 171.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 166.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 208.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 182.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 223.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 201.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 168.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 208.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 184.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 224.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 162.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 201.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 166.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 150.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 198.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 183.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 183.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 168.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 215.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 198.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 167.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 169.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 151.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 156.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 159.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 144.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 191.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 176.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 167.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 153.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 193.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 184.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 184.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 171.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 212.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 200.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:27 171.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 165.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 154.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 156.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 160.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 146.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 185.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 176.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 154.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 178.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 174.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 198.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 148.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 171.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 158.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 181.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 178.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 202.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 151.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 174.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 139.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 139.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 143.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 142.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 152.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 176.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 167.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 191.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 146.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 170.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 156.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 179.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 195.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 150.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 173.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 152.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 135.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 172.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 157.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 167.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 153.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 190.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 172.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 148.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 144.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:25 131.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 130.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 145.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 129.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 165.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 150.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 156.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 139.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 173.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 159.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 172.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 158.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 191.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 175.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 152.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 145.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 136.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 132.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 149.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 133.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 166.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 151.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 148.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 163.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 168.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 183.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 142.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 156.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 152.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 167.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 173.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 188.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 146.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 159.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 132.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 124.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 136.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 127.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 146.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 161.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 161.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 176.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 140.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 154.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 150.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 165.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 166.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 181.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 144.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 158.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 146.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 130.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 158.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 141.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 161.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 145.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 174.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 159.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:28 140.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 129.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:26 124.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 114.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 139.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 123.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 151.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 134.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 150.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 134.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 162.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 145.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 166.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 152.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 179.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 164.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:25 144.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 132.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:24 128.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 117.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 143.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 128.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 154.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 138.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 198.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 195.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 182.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 156.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 160.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 147.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 138.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 141.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 129.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 188.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 247.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 180.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 238.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 190.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 246.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 182.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 237.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 197.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 197.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 185.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 243.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 207.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 266.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 178.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 236.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 187.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 242.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 209.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 266.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 179.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 235.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 187.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 173.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 229.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 230.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 211.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 195.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 257.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 253.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 194.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 191.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 178.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 165.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 220.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 222.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 188.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 174.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 220.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 229.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 211.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 197.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 246.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 253.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 183.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 191.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 178.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 166.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 211.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 221.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 171.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 203.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 163.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 195.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 176.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 210.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 167.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 201.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 155.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 159.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 167.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 200.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 189.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 222.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 160.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 192.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 173.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 206.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 195.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 230.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 165.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 199.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 169.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 154.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 198.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 187.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 191.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 176.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 222.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 210.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 160.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 149.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 160.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 147.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 188.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 179.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 174.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 159.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 198.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 193.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 197.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 183.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 222.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 217.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:25 158.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 152.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 165.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 152.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 188.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 185.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 163.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 185.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 155.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 177.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 169.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 191.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 160.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 182.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 137.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 141.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 160.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 182.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 181.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 205.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 153.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 175.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 166.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 188.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 188.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 211.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 158.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 180.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 146.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 179.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 169.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 184.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 169.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 203.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 193.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 141.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 131.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 152.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 139.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 170.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 161.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 167.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 153.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 184.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 174.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 190.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 176.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 208.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 199.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:24 144.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 135.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 158.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 145.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 174.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 166.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 175.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 168.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 176.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 171.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 182.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 166.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 175.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 158.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 188.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 169.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:23 183.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 170.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 164.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 156.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 192.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 250.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 183.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 241.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 193.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 249.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 185.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 240.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 204.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 205.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 188.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 246.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 210.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 269.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 181.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 239.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 190.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 246.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 212.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 269.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 183.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 238.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 189.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 162.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 232.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 218.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 213.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 183.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 258.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 240.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 202.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 188.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 180.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 154.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 223.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 210.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 190.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 164.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 223.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 215.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 213.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 186.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 248.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 238.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:28 191.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 183.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 181.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 156.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 213.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 207.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 174.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 207.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 165.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 198.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 179.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 213.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 170.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 204.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 163.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 166.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 170.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 203.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 193.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 226.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 163.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 196.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 175.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 210.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 198.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 233.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 168.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 202.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 171.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 144.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 200.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 183.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 194.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 165.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 224.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 206.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 168.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 152.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 162.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 136.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 191.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 175.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 176.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 150.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 201.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 185.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 199.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 172.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 225.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 208.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 165.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:24 152.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 167.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 142.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 190.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 176.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 166.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 189.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 158.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 180.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 172.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 194.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 163.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 185.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 145.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:25 149.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 163.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 185.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 185.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 208.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 156.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 178.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 169.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 191.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 191.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 214.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 161.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 183.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 163.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 138.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 182.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 165.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 186.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 159.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 206.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 188.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:27 149.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:28 134.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 155.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 130.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 173.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 157.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 170.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 144.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 187.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 170.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:24 193.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 166.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 211.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 194.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:26 152.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:27 138.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 160.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 136.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:23 177.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 162.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 100.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 88.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 104.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 92.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 99.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:24 107.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:27 94.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:26 135.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:28 123.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:25 109.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:28 96.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 88.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:27 103.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:28 111.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:26 98.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:28 139.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:27 126.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:27 112.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:24 100.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 92.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 110.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 99.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 114.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 103.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 130.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 108.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 119.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 115.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 101.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 134.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 113.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 105.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 121.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 107.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 115.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 105.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 101.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 121.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 111.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 120.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 107.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 104.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 119.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 105.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 109.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 93.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 116.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 104.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 107.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 103.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 91.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 100.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 87.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 120.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 108.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 113.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 98.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 120.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 108.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 126.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 112.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 106.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 94.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 103.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 91.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 128.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 104.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 114.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 112.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 97.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 131.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 108.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 114.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 101.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 119.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 111.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 102.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 96.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 118.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 107.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 116.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 104.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 100.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 116.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 102.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 104.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 89.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:26 113.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 102.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 116.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 103.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 88.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 95.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 83.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 117.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 105.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 109.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 93.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 117.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 106.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 121.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 108.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 103.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 91.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 99.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 87.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 126.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 102.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 112.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 110.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 95.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 130.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 106.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 117.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 113.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 99.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 117.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 100.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 109.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 100.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 94.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 117.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 105.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 114.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 103.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 98.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 115.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 101.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 102.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 87.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:28 112.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:25 100.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 114.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 100.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 99.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 87.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 93.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 81.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 104.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 106.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 91.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:25 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 104.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 119.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 105.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 102.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 89.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 97.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 85.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 129.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 105.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 115.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 114.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 98.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 132.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 110.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 120.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 116.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 102.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 120.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 103.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 112.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 97.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 120.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 108.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 117.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 106.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:23 102.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 117.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 103.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:26 104.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 89.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 113.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:26 101.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 116.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 103.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 101.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 89.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:27 95.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 83.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 118.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 105.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 109.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:24 94.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:24 117.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 105.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:28 108.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 104.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 92.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:25 99.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 87.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 135.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 107.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 116.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 100.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 138.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 126.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 106.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 114.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 99.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 126.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 118.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 104.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 105.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 90.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 114.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 102.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:28 98.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 86.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 119.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 132.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 103.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 112.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 96.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 135.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 123.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:23 101.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 110.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:28 123.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 115.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:24 101.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 101.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 86.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 110.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 98.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 93.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 81.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 104.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 130.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:25 101.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 109.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 94.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 133.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 99.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 108.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 93.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:27 100.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:25 98.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 84.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 107.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 95.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 91.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 79.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 103.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 134.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 103.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:28 113.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 97.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 136.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 125.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 102.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 111.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 96.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:25 124.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 116.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 101.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:26 86.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 110.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 98.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 93.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 81.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 116.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:26 105.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 115.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 104.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 119.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 108.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 112.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 102.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 116.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 106.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 111.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:28 100.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 115.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:24 104.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 112.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 101.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 116.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 105.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 215.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 86.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 94.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 98.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 82.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 204.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:27 84.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:24 91.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:26 89.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 80.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:34 204.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 190.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:26 84.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 71.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:27 92.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 79.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:39 89.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 76.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 82.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 68.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:23 138.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 113.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 123.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 121.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 106.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 141.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 117.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 129.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 123.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 110.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 129.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 120.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 111.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 105.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 128.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 116.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 125.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 113.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 127.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 97.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 130.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 118.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 125.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 97.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 91.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 115.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 101.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 132.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 120.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 130.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 117.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 112.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 99.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 95.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 135.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 109.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 118.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 102.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 138.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 113.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 124.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 105.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 108.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 101.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 105.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 124.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 93.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 127.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 115.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 121.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 108.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 94.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 87.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 124.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 113.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 97.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 129.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 117.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 96.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 90.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 133.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 106.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 99.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 137.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 122.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 119.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 103.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 105.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 114.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 98.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 124.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 109.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 109.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 102.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 122.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 108.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 126.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 114.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 105.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 104.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 92.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 96.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 84.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 123.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 111.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 95.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 127.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 115.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 123.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 107.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 95.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 88.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 137.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 110.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 140.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 115.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 127.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 123.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 128.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 109.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 118.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 111.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 103.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 127.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 124.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 113.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 110.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 94.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 128.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 116.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 122.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 95.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 88.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 126.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 101.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 129.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 118.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 127.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 114.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 97.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 94.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 144.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 112.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 122.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 142.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 129.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 141.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 108.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 118.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 100.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 139.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 127.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 139.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 105.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 116.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 98.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 137.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 143.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 109.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 120.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 102.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 140.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 128.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 100.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 88.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 104.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 92.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 99.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:33 107.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:39 94.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:38 135.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:30 123.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:34 109.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:36 96.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 88.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 103.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:30 111.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:36 98.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:34 139.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:38 126.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:40 113.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:37 100.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 92.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 96.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 84.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 99.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 87.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 97.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 85.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 100.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 88.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 108.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 110.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 98.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 125.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 135.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 118.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 130.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 128.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 140.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 119.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 115.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 131.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 118.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 125.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 135.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 108.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 120.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 118.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 114.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 106.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 100.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 125.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 111.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 138.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 126.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 91.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 101.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 121.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 109.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 104.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 129.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 142.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 130.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 120.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 129.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 110.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 111.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 124.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 122.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 134.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 113.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 108.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 99.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 109.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 119.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 129.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 102.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 111.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 100.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 94.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 131.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 118.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 97.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 85.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 94.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 114.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 103.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 98.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 122.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 136.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 123.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 100.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 88.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 110.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 121.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 119.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 131.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 110.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 105.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 96.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 109.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 126.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 99.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 108.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 97.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 91.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 128.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 115.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 82.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 91.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 111.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 107.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 100.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 119.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 133.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 120.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 97.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 85.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 95.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 128.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 109.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 110.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 123.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 133.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 112.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 107.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 123.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 98.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 109.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 111.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 128.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 101.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 111.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 99.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 93.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 129.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 96.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 84.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 104.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 92.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 114.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 108.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 101.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 97.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 120.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 135.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 120.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 99.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 87.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 97.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 121.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 107.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 124.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 110.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 112.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 101.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 116.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 103.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 109.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 98.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 112.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 100.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 111.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 99.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 114.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 102.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 147.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 102.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 111.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 119.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 98.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 135.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 99.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 108.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 96.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 134.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 121.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 100.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 86.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 109.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 95.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:37 108.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 94.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 95.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 83.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 135.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 129.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 140.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 122.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 122.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 137.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 133.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 144.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 124.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 125.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 123.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 135.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 111.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 130.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 139.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 113.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 124.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 134.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 122.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 118.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 129.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 143.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 109.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 97.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 135.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 121.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 133.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 118.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 146.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 133.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 111.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 99.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 124.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 128.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 122.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 134.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 132.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 126.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 139.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 118.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 116.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 120.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 129.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 124.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 134.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 126.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 114.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 111.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 122.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 105.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 136.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 121.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 90.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 97.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 127.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 117.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 114.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 105.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 126.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 112.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 140.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 127.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 93.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 102.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 125.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 119.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 131.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 112.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 129.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 123.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 136.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 115.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 113.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 101.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 111.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 116.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 121.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 131.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 104.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 115.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 123.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 110.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 108.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 133.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 118.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 87.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 110.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 94.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 124.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 111.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 102.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 123.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 137.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 123.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 102.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 91.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 114.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 99.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 128.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 121.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 133.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 115.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 115.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 132.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 126.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 138.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 118.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 119.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 115.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 119.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 129.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 104.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 113.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 134.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 107.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 117.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 125.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 111.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 101.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 121.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 135.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 120.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 102.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 90.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 126.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 116.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 113.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 105.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 125.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 139.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 126.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 93.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 102.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 97.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 84.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 100.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 87.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 97.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 85.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 100.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 88.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 124.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 131.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 119.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 119.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 122.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 129.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 118.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 109.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 96.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 115.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 103.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 113.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 101.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 104.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 94.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 105.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 92.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 108.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 95.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 106.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 93.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 109.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 96.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 135.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 130.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 121.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 116.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 121.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 137.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 131.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 124.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 121.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 115.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 105.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 122.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 118.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 110.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 105.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 109.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 96.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 125.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 121.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 113.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 108.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 112.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 99.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 98.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 111.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 101.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 118.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 106.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 100.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 104.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 91.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 120.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 108.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 95.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 138.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 148.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 134.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:32 118.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 121.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 136.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 145.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 133.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 124.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 110.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 158.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 119.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:40 118.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 116.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:32 105.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 104.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 118.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 107.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 136.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 122.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 139.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 125.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 124.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 111.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 127.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 114.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 107.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 122.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 110.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 131.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 119.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 132.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 137.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 125.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 126.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 138.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 113.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 108.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 123.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 111.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 125.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 101.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 129.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 116.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 104.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 103.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 106.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 121.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 108.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 109.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 96.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 124.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 112.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 113.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 100.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 171.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 162.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 172.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 163.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 167.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 179.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 160.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 168.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 183.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 161.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 171.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 155.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 190.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 177.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 156.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 144.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 173.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 157.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 195.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 180.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 157.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 144.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 155.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 146.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 160.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 151.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 151.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 164.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 144.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 156.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 171.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 149.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 156.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 139.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 175.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 161.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 140.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 127.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 161.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 145.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 182.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 166.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 145.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 133.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 148.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 140.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 154.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 145.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 144.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 158.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 137.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 150.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 165.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 143.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 149.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 133.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 169.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 154.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 134.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 155.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 139.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 175.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 161.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 139.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 127.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 156.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 150.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 141.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 153.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 163.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 147.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 143.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 127.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 154.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 138.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:39 135.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 122.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 131.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 124.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 175.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 166.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 178.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 169.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 171.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 184.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 164.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 175.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 187.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 167.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 175.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 159.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 195.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 180.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 164.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 147.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 177.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 162.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 199.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 184.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 166.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 147.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 159.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 150.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 165.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 155.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 155.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 169.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 148.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 161.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 176.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 153.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 160.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 144.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 180.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 163.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 149.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 130.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 165.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 149.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 186.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 170.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 154.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 136.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 153.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 144.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 159.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 149.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 149.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 162.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 142.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 155.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 170.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 147.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 153.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 137.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 173.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 156.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 143.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 159.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 143.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 179.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 163.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 148.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 130.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 94.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 82.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 98.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 86.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 94.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:32 110.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:36 87.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:31 138.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:38 116.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:35 112.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:34 89.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 82.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 98.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:39 110.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:33 91.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:37 139.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:38 120.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:31 113.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:30 93.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 86.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 105.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 93.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 109.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 97.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 117.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 108.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 100.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 123.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 108.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 112.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 100.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 103.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 108.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 94.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 103.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 91.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 111.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 98.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 83.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 86.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 112.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 99.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 95.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 115.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 102.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 87.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 102.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 90.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 114.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 99.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 110.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 96.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 120.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 97.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 95.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 100.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 92.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 86.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 108.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 95.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 112.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 99.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 96.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 81.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 93.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 82.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 97.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 103.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 91.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 112.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 99.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 117.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 99.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 84.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 97.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 85.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 97.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 108.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 103.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 94.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 119.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 102.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 106.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 103.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 96.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 96.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 93.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 109.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 99.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 97.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 90.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 96.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 84.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 106.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 94.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 97.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 94.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 79.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 91.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 79.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 95.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 88.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 110.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 98.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 98.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 82.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 83.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 111.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 107.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 96.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 122.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 105.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 110.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 107.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 99.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 109.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 99.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 96.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 112.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 104.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 114.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 92.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 98.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 86.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 108.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 95.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 99.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 96.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 81.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 93.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 82.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 97.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 91.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 112.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 99.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 117.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 104.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 100.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 85.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 97.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 86.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 102.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 112.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 99.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 127.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 112.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 101.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 113.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 98.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 117.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 93.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 100.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 88.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 110.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 98.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 96.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 84.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 110.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 98.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 98.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 95.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 97.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 108.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 93.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 104.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 91.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 95.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 83.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 106.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 94.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 91.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 80.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 107.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 95.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 117.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 96.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 105.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 92.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 122.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 108.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 95.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 106.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 91.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 113.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 89.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 93.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 81.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 104.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 92.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 89.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 78.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 94.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 99.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 108.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 111.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 97.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 109.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 94.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 116.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 91.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 95.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 83.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 106.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 94.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 91.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 80.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 96.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 109.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 98.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 113.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 102.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 106.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 95.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 110.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 99.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 104.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 94.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 108.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 98.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 106.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 95.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 110.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 99.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 209.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 82.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 90.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 95.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 81.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 201.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 81.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 88.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 87.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 80.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 200.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 185.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 83.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 70.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 89.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 76.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 85.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 74.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 81.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 68.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 124.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 119.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 130.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 108.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 107.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 121.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 122.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 109.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 102.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 107.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 94.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 125.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 112.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 104.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 89.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 89.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 107.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 126.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 114.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 125.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 92.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 93.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 122.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 104.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 115.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 128.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 108.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 120.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 103.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 113.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 99.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 118.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 113.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 99.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 102.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 90.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 122.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 109.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 116.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 101.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 86.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 97.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 85.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 116.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 104.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 94.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 111.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 109.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 89.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 102.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 89.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 102.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 110.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 106.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 113.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 102.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 111.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 97.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 116.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 105.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 105.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 101.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 98.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 100.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 88.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 120.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 108.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 100.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 84.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 95.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 83.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 102.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 92.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 122.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 119.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 103.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 88.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 99.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 87.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 114.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 102.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 129.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 111.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 116.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 106.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 105.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 101.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 120.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 108.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 114.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 100.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 103.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 91.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 122.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 110.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 118.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 105.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 103.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 90.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 98.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 86.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 105.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 108.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 124.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 111.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 123.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 110.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 93.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 104.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 92.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 130.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 107.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 119.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 103.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 130.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 127.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 103.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 115.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 98.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 127.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 101.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 113.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 96.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 129.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 105.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 117.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 100.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 131.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 117.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:39 110.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:36 87.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:38 138.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:36 116.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:36 112.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:36 89.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:37 111.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:38 91.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:33 139.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:36 120.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:36 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:33 93.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 98.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 86.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 102.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 90.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 98.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 86.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 102.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 90.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 103.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 91.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 107.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 94.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 102.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 92.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 107.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 96.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 131.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 132.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 142.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 117.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 126.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 137.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 136.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 145.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 122.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 129.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 120.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 129.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 139.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 105.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 124.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 126.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 133.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 142.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 111.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 128.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 118.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 106.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 131.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 114.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 141.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 126.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:33 104.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 92.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 120.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 108.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:34 122.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 112.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 133.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 117.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 144.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 129.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:33 111.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 97.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 121.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 111.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 121.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 123.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 133.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 106.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 117.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 124.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 127.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 138.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 109.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 110.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 121.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 130.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 95.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 115.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 125.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 135.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 97.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 119.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:34 108.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 97.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 121.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 105.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 131.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 117.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 94.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 82.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 111.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 100.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 111.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 100.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 125.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 109.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 136.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 122.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 96.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 84.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 114.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 103.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 116.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 119.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 129.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 101.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 113.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 119.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 123.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 135.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 104.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 117.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 105.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 117.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 126.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 90.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 111.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 108.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 121.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 131.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 93.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 116.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 103.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 92.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 117.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 101.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 127.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 113.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 89.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 77.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 107.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 96.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:30 106.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 95.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 121.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 106.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 133.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 119.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 91.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 79.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 111.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 100.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 142.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 135.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 146.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 129.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 148.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 139.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 149.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 129.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 132.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 131.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 133.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 142.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 113.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 127.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 137.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 136.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 146.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 118.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 131.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 129.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 116.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 133.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 143.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 128.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 112.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 98.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 123.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 110.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:31 132.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 122.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 136.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 118.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 147.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 130.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 119.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 103.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 124.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 112.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 132.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 126.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 137.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 114.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 120.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 135.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 130.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 142.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 116.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 124.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 124.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 133.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 103.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 118.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 124.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 128.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 138.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 105.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 122.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 119.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 106.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 124.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 106.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 135.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 118.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:37 101.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 88.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 113.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 101.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 121.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 109.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 128.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 111.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 139.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:35 103.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 90.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 117.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 105.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 127.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 122.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 133.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 109.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 116.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 130.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 127.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 138.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 112.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 120.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 116.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 120.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 129.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 98.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 114.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 119.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 125.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 135.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 100.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 119.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:30 113.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 101.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 120.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 102.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 131.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 114.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:37 96.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 83.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 97.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:33 116.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 104.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 124.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 107.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 135.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 120.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 99.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 86.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 114.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 101.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 109.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 115.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 103.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 98.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 104.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 91.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 93.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 99.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 87.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 126.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 139.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 137.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 151.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 119.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 132.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 129.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 141.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 141.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 154.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 122.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 134.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 107.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 109.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 125.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 138.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 134.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 148.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 118.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 131.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 128.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 140.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 138.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 151.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 121.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 133.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 124.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 111.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 138.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 124.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 137.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 122.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 151.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 135.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 102.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 119.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 106.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 132.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 118.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 127.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 115.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 138.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 126.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 140.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 125.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 153.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 138.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:37 113.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 103.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 122.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 109.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 133.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 120.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 121.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 128.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 132.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 140.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 114.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 121.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 125.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 131.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 136.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 144.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 118.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 124.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 97.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 99.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 120.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 127.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 129.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 137.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 114.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 120.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 124.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 130.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 133.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 141.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 117.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 123.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 119.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 107.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 126.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 113.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 132.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 117.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 139.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 124.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:34 102.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 91.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 114.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 101.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 120.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 108.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 123.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 110.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 129.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 116.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 136.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 121.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 143.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 128.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 104.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 93.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 117.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 104.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 123.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 110.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 119.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 123.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 130.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 135.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 112.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 116.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 122.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 127.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 134.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 139.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 115.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 119.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 92.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 94.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 118.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 122.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 127.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 132.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 111.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 115.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 121.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 126.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 131.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 136.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 114.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 118.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 117.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 104.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 121.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 109.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 129.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 114.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 134.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 120.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:33 97.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 86.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 112.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 99.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 115.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 103.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 121.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 108.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 125.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 112.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 134.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 119.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 139.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 124.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:39 99.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 89.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 115.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 102.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 119.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 106.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 113.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 120.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 109.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 108.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 110.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 117.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 107.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 112.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 98.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 120.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 105.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 115.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 103.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 108.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 96.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 129.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 143.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 140.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 155.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 122.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 135.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 132.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 145.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 144.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 157.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 125.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 137.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 115.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 116.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 128.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 141.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 137.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 152.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 121.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 134.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 131.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 143.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 141.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 154.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 124.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 136.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 127.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 112.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 141.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 125.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 140.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 122.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 154.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 136.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:39 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 108.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 121.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 106.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 135.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 119.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 130.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 115.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 141.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 126.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 143.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 126.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 156.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 138.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:32 120.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 108.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 109.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 135.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 120.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 124.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 131.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 135.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 143.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 117.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 124.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 128.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 135.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 139.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 147.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 121.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 127.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 104.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 107.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 123.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 130.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 133.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 140.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 116.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 123.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 127.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 133.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 137.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 144.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 120.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 126.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 122.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 107.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 129.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 135.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 118.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 142.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:35 109.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 97.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 117.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 102.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 123.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 125.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 111.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 132.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 117.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 139.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 122.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 146.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 128.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 111.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 99.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 120.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 105.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 126.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 122.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 127.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 133.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 138.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 115.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 119.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 126.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 130.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 137.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 143.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 118.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 122.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 99.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 102.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 125.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 130.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 135.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 114.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 118.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 124.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 129.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 134.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 140.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 117.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 121.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 120.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 105.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 124.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 109.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 132.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 115.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 137.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 120.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 92.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 114.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 99.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 118.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 103.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 123.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 109.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 127.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 113.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 136.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 142.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 124.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:33 107.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 95.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 118.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 103.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 121.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 106.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 113.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 119.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 104.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 105.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 90.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 92.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 97.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 83.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 135.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 151.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 153.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 171.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 129.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 145.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 139.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 153.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 158.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 173.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 133.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 146.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:33 108.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 112.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:39 111.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 114.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 133.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 149.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 147.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 164.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 127.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 144.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 137.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 151.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 152.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 167.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 131.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 145.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 133.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 120.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 150.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 136.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 150.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 135.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 167.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 155.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:35 116.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:34 118.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:37 103.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 106.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 127.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 113.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 143.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 130.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 137.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 124.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 151.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 137.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 154.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 139.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 169.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 157.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:40 119.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:30 119.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:32 105.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 108.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 130.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 117.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 144.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 130.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 130.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 138.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 148.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 157.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 123.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 132.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 134.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 142.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 153.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 162.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 128.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 135.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:39 101.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 97.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:29 104.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 99.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 127.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 136.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 142.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 151.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 122.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 130.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 132.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 140.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 147.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 156.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 126.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 134.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 127.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 114.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 136.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 123.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 144.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 129.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 153.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 139.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:36 108.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 104.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:31 95.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 92.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 121.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 108.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 129.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 116.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 132.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 119.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 140.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 127.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 149.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 135.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 158.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 146.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:33 111.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:32 105.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:33 98.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 94.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 125.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 112.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 133.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 120.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 127.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 132.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 145.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 152.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 120.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 126.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 131.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 137.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 150.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 157.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 125.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 130.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:35 96.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 90.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:40 99.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 93.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 124.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 130.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 139.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 145.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 119.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 125.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 129.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 135.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 144.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 151.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 123.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 129.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 125.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 111.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 130.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 117.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 141.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 126.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 147.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 133.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:38 103.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 96.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:31 90.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 84.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 118.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 105.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 123.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 111.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 129.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 116.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 135.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 122.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 146.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 132.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 153.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 139.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:35 105.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:34 98.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:33 93.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 87.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 123.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 109.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 128.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 115.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 122.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 130.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 118.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:30 106.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 110.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 119.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 127.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 119.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 106.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 129.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 114.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:38 117.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 120.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:36 100.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 105.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 115.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 104.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 138.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 155.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 157.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 174.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 132.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 148.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 142.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 156.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 161.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 177.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 136.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 149.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:35 115.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 119.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:33 117.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 122.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 136.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 153.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 150.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 168.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 130.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 147.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 140.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 154.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 155.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 170.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 134.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 148.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 136.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 120.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 152.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 136.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 152.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 135.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 170.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 154.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:37 123.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:30 125.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:33 107.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 110.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 129.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 114.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 145.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 130.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 140.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 124.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 154.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 138.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 157.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 140.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 172.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 156.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:39 125.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 126.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:36 110.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 113.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 133.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 118.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 146.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 131.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 133.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 141.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 151.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 161.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 126.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 135.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 137.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 145.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 156.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 166.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 131.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 138.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:40 108.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 105.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:40 110.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 107.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 131.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 139.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 145.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 154.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 125.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 133.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 135.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 143.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 150.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 159.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 129.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 137.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 130.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 115.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 139.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 123.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 147.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 129.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 156.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 138.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:36 114.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 111.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:40 101.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:32 95.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 124.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 108.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 132.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 116.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 135.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 120.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 143.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 127.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 152.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 135.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 161.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 146.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:32 117.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:38 103.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 97.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 128.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 113.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 135.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 121.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 130.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 136.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 148.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 155.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 123.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 129.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 134.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 140.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 153.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 161.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 128.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 133.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:35 103.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 98.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:29 106.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 100.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 128.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 134.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 142.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 149.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 122.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 128.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 132.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 138.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 147.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 154.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 126.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 132.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 127.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 112.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 133.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 118.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 144.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 127.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 151.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 133.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:35 109.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:34 103.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:36 96.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 88.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 121.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 106.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 126.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 111.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 132.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 117.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 138.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 122.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 149.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 132.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 156.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 139.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 16-Jul-2026 09:37 112.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 106.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 16-Jul-2026 09:36 99.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 91.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 125.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 110.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 130.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 115.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 129.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 134.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 110.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 115.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 103.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 94.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 152.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 175.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 144.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 167.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 158.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 176.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 150.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 167.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 128.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 126.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 149.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 172.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 171.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 194.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 142.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 165.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 155.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 173.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 177.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 196.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 147.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 165.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 151.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 136.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 174.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 159.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 173.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 158.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 198.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 182.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:40 134.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 122.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 141.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 128.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 165.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 151.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 156.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 141.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 174.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 160.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 179.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 164.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 199.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 183.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:29 130.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 121.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 146.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 134.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 165.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 152.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 145.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 158.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 137.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 149.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 151.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 163.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 143.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 155.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 108.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:30 109.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 142.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 154.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 163.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 176.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 135.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 147.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 148.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 160.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 170.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 183.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 141.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 152.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 143.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 129.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 156.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 141.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 166.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 151.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 179.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 164.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:35 115.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 102.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 134.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 121.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 146.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 133.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 149.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 135.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 161.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 147.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 172.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 158.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 185.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 170.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:35 116.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 103.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 140.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 127.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 151.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 139.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 141.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 150.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 133.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 142.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 148.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 157.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 139.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 148.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 99.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 102.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 138.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 147.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 160.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 169.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 131.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 140.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 145.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 154.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 167.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 176.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 137.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 146.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 140.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 125.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 148.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 134.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 162.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 147.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 172.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 157.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 106.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:29 93.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 131.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 117.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 139.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 126.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 146.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 132.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 155.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 140.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 169.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 154.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 179.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 164.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:38 108.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:36 96.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 137.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 124.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 145.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 132.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 138.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 131.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 124.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 133.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 144.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 128.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 140.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 121.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 150.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 132.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:37 133.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 118.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 128.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 118.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 156.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 178.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 147.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 170.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 161.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 180.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 152.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 170.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:39 135.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:34 134.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 152.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 175.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 174.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 198.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 145.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 168.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 158.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 176.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 180.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 199.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 150.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 168.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 153.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 132.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 177.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 147.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 176.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 153.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 200.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 170.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:37 141.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 129.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 144.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 124.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 167.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 140.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 159.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 137.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 177.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 151.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 182.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 159.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 202.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 174.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 137.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:37 131.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 149.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 129.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 167.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 143.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 148.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 161.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 140.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 152.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 155.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 167.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 146.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 158.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:40 116.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 117.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 145.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 158.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 166.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 180.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 138.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 150.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 151.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 163.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 173.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 187.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 144.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 155.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 146.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 125.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 159.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 131.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 169.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 146.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 182.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 153.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:36 122.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 108.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 137.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 117.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 149.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 123.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 152.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 131.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 164.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 138.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 175.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 153.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 188.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 161.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:37 124.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:33 110.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 143.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 123.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 154.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 129.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 145.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 154.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 136.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 145.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 151.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 160.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 142.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 151.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 107.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:35 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 141.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 150.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 163.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 173.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 134.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 143.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 148.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 157.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 170.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 180.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 140.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 149.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 142.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 121.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 151.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 125.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 165.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 142.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 175.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 147.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:35 113.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:31 99.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 133.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 113.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 142.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 117.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 149.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 127.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 157.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 132.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 172.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 149.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 182.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 154.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 16-Jul-2026 09:31 116.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 16-Jul-2026 09:38 102.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 139.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 119.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 147.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 123.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 101.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 89.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 105.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 93.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 101.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:36 109.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:32 96.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:40 137.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:39 124.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:37 111.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:33 98.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 89.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 105.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:37 113.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:33 99.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:38 141.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:37 128.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:38 115.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:36 102.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 93.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 112.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 100.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 116.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 104.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 133.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 110.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 103.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 136.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 114.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 119.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 107.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 108.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 117.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 106.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 123.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 122.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 106.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 125.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 113.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 96.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 118.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 105.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 119.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 107.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 94.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 89.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 127.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 116.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 100.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 122.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 109.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 124.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 110.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 97.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 105.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 93.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 130.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 105.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 115.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 98.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 133.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 110.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 120.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 116.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 121.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 104.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 120.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 102.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 122.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 110.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 92.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 115.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 103.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 104.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 91.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 97.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 85.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 124.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 112.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 119.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 107.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 120.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 107.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 94.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 101.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 88.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 113.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 96.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 132.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 118.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 115.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 120.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 102.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 95.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 119.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 99.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 120.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 105.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 90.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 113.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 101.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 113.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 101.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 90.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 94.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 82.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 123.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 110.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 94.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 117.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 105.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 106.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 93.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 86.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 132.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 99.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 134.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 118.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 123.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 113.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 98.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 122.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 119.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 108.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 122.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 110.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 107.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 92.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 115.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 102.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 115.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 92.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 97.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 85.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 97.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 118.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 106.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 121.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 109.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 108.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 96.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 89.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 137.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 108.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 117.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 102.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 140.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 129.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 107.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 115.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 100.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 127.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 123.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 109.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 94.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 115.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 103.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 99.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 87.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 125.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 114.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 134.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 104.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 113.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 97.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 137.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 126.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 103.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 111.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 96.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 124.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 121.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 109.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 104.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 89.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 110.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 99.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 95.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 83.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 123.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 111.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 133.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 102.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 111.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 95.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 135.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 124.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 100.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 109.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 94.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 123.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 107.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 102.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 87.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 108.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 97.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 93.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 81.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 121.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 136.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 105.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 114.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 98.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 138.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 103.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 112.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 97.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 126.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 121.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 104.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 89.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 110.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 99.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 95.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 83.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 123.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 117.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 105.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 121.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 109.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 114.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 103.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 118.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 107.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 112.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 101.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 116.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 105.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 113.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 102.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 117.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 106.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 140.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 114.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 124.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 123.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 107.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 143.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 119.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 129.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 125.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 132.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 113.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 106.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 131.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 117.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 115.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 110.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 132.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 120.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 116.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 132.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 119.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 112.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 100.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 92.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 134.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 122.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 120.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 133.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 121.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 129.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 103.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 109.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 96.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 137.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 110.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 120.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 120.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 103.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 140.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 114.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 125.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 122.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 107.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 129.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 109.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 117.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 102.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 128.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 113.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 123.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 113.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 129.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 117.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 96.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 129.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 116.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 120.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 97.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 88.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 131.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 119.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 100.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 130.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 118.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 125.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 113.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 100.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 105.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 92.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 136.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 118.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 119.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 139.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 123.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 121.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 127.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 115.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 100.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 126.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 120.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 111.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 104.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 94.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 127.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 115.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 117.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 106.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 108.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 96.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 98.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 86.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 130.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 118.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 98.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 129.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 117.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 122.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 111.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 111.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 98.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 102.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 90.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 140.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 122.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 123.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 142.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 117.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 128.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 109.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 131.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 110.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 119.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 113.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 130.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 125.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 114.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 131.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 118.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 113.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 97.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 129.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 117.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 110.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 111.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 99.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 89.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 133.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 121.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 131.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 119.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 127.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 101.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 108.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 96.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 147.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 113.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 123.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 105.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 147.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 137.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 144.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 109.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 119.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 101.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 144.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 134.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 142.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 107.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 117.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 99.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 143.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 133.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 146.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 111.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 121.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 103.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 146.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 135.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 101.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:32 109.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:30 96.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:35 137.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:30 124.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:37 111.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:34 98.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 89.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 105.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:38 113.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:35 100.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:36 141.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:39 128.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:30 115.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:35 102.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 93.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 97.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 85.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 100.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 88.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 98.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 86.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 101.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 88.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 109.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 96.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 111.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 99.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 129.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 137.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 120.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 132.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 129.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 142.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 121.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 122.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 124.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 133.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 118.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 127.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 137.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 110.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 121.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 119.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 115.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 107.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 101.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 130.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 139.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 127.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 109.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 90.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 116.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 103.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 122.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 118.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 109.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 104.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 134.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 144.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 131.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 93.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 122.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 119.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 131.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 123.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 136.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 115.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 116.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 116.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 126.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 101.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 131.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 104.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 115.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 112.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 107.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 101.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 95.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 122.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 132.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 84.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 97.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 115.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 103.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 137.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 105.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 87.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 119.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 128.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 123.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 120.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 133.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 107.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 123.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 98.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 111.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 128.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 101.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 112.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 109.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 104.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 97.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 92.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 119.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 129.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 99.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 81.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 94.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 112.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 108.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 100.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 96.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 109.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 134.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 121.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 102.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 84.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 118.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 130.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 111.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 125.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 122.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 135.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 114.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 116.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 125.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 100.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 110.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 113.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 120.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 130.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 103.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 112.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 106.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 99.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 94.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 121.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 107.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 130.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 101.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 82.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 107.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 95.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 115.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 109.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 102.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 125.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 136.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 123.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 104.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 86.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 111.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 122.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 108.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 125.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 110.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 114.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 101.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 116.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 104.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 110.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 98.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 113.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 101.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 112.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 100.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 115.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 103.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 137.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 131.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 142.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 124.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 124.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 139.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 134.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 146.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 126.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 127.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 124.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 128.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 137.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 114.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 122.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 132.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 141.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 125.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 135.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 123.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 119.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 109.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 134.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 120.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 144.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 132.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 95.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 124.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 136.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 121.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 112.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 138.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 123.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 149.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 136.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 98.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 127.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 110.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 130.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 136.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 117.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 133.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 128.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 141.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 121.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 117.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 131.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 115.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 121.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 126.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 136.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 119.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 127.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 112.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 127.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 137.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 124.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 107.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 89.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 116.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 128.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 115.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 131.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 117.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 142.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 129.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 110.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 92.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 120.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 120.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 133.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 114.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 130.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 125.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 138.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 118.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 118.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 118.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 128.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 104.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 112.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 118.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 123.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 133.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 124.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 111.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 109.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 123.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 134.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 104.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 86.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 113.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 97.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 125.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 112.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 103.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 128.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 139.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 89.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 117.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 102.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 129.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 123.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 135.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 133.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 140.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 120.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 120.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 120.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 131.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 121.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 125.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 136.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 110.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 119.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 126.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 114.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 111.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 102.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 125.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 111.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 136.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 123.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 106.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 88.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 101.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 127.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 114.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 130.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 141.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 128.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 109.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 91.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 105.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 106.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 93.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 109.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 96.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 107.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 94.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 110.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 97.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 124.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 125.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 116.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 135.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 131.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 122.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 126.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 138.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 132.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 125.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 118.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 128.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 106.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 103.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 115.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 123.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 119.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 111.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 106.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 114.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 126.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 122.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 114.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 109.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 97.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 99.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 111.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 102.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 119.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 107.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 101.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 109.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 90.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 121.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 118.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 109.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 112.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 93.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 137.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 123.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 140.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 126.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 125.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 112.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 128.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 115.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 120.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 107.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 123.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 110.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 131.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 121.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 133.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 138.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 131.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 140.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 126.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 133.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 109.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 123.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 119.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 99.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 130.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 117.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 122.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 115.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 104.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 122.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 95.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 125.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 117.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 98.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 172.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 163.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 174.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 165.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 168.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 183.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 161.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 170.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 186.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 162.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 179.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 165.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 195.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 181.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 161.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 150.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 182.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 168.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 198.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 183.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 161.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 150.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 156.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 147.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 162.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 153.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 152.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 168.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 145.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 158.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 175.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 150.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 164.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 150.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 179.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 164.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 145.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 133.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 170.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 155.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 185.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 170.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 150.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 138.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 150.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 141.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 156.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 147.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 146.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 161.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 139.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 152.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 168.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 145.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 157.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 143.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 173.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 157.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 139.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 126.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 163.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 149.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 179.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 164.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 144.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 133.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 176.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 167.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 180.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 171.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 173.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 187.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 165.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 176.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 191.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 168.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 184.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 169.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 199.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 185.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 170.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 153.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 186.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 172.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 202.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 188.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 172.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 153.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 161.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 152.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 166.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 157.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 157.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 173.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 150.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 163.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 179.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 155.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 169.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 154.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 184.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 168.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 155.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 136.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 174.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 159.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 190.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 175.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 159.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 142.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 155.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 145.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 160.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 151.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 151.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 166.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 143.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 157.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 173.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 149.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 162.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 147.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 177.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 162.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 148.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 130.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 167.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 153.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 184.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 169.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 153.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 136.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 102.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 82.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 104.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 86.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 103.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:36 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:35 89.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:31 129.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:38 117.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:31 102.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:32 91.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 82.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 104.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:38 104.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:34 92.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:34 133.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:38 121.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:32 106.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:36 94.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 86.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 113.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 93.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 114.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 97.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 115.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 109.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 101.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 125.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 109.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 120.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 112.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 110.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 114.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 101.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 118.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 104.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 104.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 111.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 98.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 104.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 92.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 119.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 98.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 106.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 101.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 86.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 99.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 87.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 104.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 96.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 120.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 102.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 124.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 111.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 104.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 89.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 103.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 91.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 100.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 111.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 107.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 97.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 122.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 105.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 101.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 107.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 100.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 109.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 99.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 114.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 101.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 96.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 88.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 116.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 95.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 102.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 83.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 95.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 83.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 113.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 101.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 104.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 92.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 117.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 99.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 120.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 107.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 86.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 87.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 114.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 109.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 105.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 95.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 120.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 108.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 99.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 105.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 97.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 107.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 97.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 94.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 111.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 102.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 100.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 98.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 94.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 86.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 114.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 94.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 96.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 81.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 93.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 81.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 99.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 102.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 90.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 116.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 98.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 100.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 85.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 96.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 85.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 101.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 109.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 98.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 123.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 106.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 117.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 102.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 109.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 100.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 110.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 101.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 105.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 103.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 108.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 96.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 100.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 88.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 116.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 95.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 99.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 83.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 95.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 83.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 101.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 104.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 93.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 117.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 99.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 120.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 87.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 87.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 123.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 103.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 113.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 100.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 102.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 114.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 99.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 109.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 97.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 101.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 90.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 114.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 102.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 98.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 86.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 114.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 99.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 109.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 96.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 126.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 111.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 98.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 110.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 95.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 116.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 106.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 95.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 97.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 85.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 109.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 97.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 93.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 82.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 111.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 99.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 119.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 97.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 107.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 93.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 109.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 96.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 107.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 93.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 93.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 95.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 83.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 107.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 95.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 91.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 79.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 98.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 122.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 100.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 110.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 96.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 127.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 113.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 99.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 110.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 95.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 107.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 95.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 97.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 85.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 109.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 97.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 93.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 82.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 100.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 118.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 98.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 120.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 102.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 115.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 95.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 117.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 99.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 113.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 94.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 115.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 98.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 115.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 95.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 116.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 99.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 126.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 110.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 121.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 132.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 114.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 125.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 118.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 109.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 117.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 109.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 119.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 108.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 122.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 124.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 110.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 118.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 106.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 96.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 130.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 112.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 91.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 91.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 123.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 132.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 114.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 128.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 110.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 94.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 95.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 123.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 105.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 113.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 101.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 129.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 121.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 114.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 115.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 105.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 109.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 120.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 108.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 104.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 115.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 103.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 92.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 127.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 109.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 119.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 107.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 103.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 88.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 87.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 120.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 96.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 129.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 111.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 124.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 91.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 90.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 122.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 114.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 111.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 119.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 112.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 102.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 113.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 104.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 118.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 106.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 117.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 101.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 89.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 126.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 108.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 105.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 102.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 87.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 96.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 84.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 119.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 94.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 127.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 109.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 122.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 110.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 105.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 90.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 101.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 88.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 107.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 131.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 124.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 116.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 102.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 111.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 116.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 104.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 105.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 93.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 128.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 110.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 121.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 108.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 105.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 92.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 100.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 88.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 121.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 109.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 110.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 100.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 129.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 111.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 126.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 113.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 108.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 95.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 106.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 94.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 132.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 108.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 121.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 104.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 133.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 123.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 129.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 104.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 116.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 100.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 130.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 120.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 127.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 102.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 114.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 97.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 128.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 119.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 131.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 106.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 118.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 101.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 133.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 121.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:39 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:40 89.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:37 129.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:33 118.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:39 102.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:39 91.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:34 104.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:36 92.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:30 133.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:31 121.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:37 106.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:32 94.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 105.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 94.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 109.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 98.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 105.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:31 120.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:31 108.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:34 148.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:35 136.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:39 122.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:31 110.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 94.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 109.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:29 124.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:40 111.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:40 152.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:40 140.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:32 126.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:36 114.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 98.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 116.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 105.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 120.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 109.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 131.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 115.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 134.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 112.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 123.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 123.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 105.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 100.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 121.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 120.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 105.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 107.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 92.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 122.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 110.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 91.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 87.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 118.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 106.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 111.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 96.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 126.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 114.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 122.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 110.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 91.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 128.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 114.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 112.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 97.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 132.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 101.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 102.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 100.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 114.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 88.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 119.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 107.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 112.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 99.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 88.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 95.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 83.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 115.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 103.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 107.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 92.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 123.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 111.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 118.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 91.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 87.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 127.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 101.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 112.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 110.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 95.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 130.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 106.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 118.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 109.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 100.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 94.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 117.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 104.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 114.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 103.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 98.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 101.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 86.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 117.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 106.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 110.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 99.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 98.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 86.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 92.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 81.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 102.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 104.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 90.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 121.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 110.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 101.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 89.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 96.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 84.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 130.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 104.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 115.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 114.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 133.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 120.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 116.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 102.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 122.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 103.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 97.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 120.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 117.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 106.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 101.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 102.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 88.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 119.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 107.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 101.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 100.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 88.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 95.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 83.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 104.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 93.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 123.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 111.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 118.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 103.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 91.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 99.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 87.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 135.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 107.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 116.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 100.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 138.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 127.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 105.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 114.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 99.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 115.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 104.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 104.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 89.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 111.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 100.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 97.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 85.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 118.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 133.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 102.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 112.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 96.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 135.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 124.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 101.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 110.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 95.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 122.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 101.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 99.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 85.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 107.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 96.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 92.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 81.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 103.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 131.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 100.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 109.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 93.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 134.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 123.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 99.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 108.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 93.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 111.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 100.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 97.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 83.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 105.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 94.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 90.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 79.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 102.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 134.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 103.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 112.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 96.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 137.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 102.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 111.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 95.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 113.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 101.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 99.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 85.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 107.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 96.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 92.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 81.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 116.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 104.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 121.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 110.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 125.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 114.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 118.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 107.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 122.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 111.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 116.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 106.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 120.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 110.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 118.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 107.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 122.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 111.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 138.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 113.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 123.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 142.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 117.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 128.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 123.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 109.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 130.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 111.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 120.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 105.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 129.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 125.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 124.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 112.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 111.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 96.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 136.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 124.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 110.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 107.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 96.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 91.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 113.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 100.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 138.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 125.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 126.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 111.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 98.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 107.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 94.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 136.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 119.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 118.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 139.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 127.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 116.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 111.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 121.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 104.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 109.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 106.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 92.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 133.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 121.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 117.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 104.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 93.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 98.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 86.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 123.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 110.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 96.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 134.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 123.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 122.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 95.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 90.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 134.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 106.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 99.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 137.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 122.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 119.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 103.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 125.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 105.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 114.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 125.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 119.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 109.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 120.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 108.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 104.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 89.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 131.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 120.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 103.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 103.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 91.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 84.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 121.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 109.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 94.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 133.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 121.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 120.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 94.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 88.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 138.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 110.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 141.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 127.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 123.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 129.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 111.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 124.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 113.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 107.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 123.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 111.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 108.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 93.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 133.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 122.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 118.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 106.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 94.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 88.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 124.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 112.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 113.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 135.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 123.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 97.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 93.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 145.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 111.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 122.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 104.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 139.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 129.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 142.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 107.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 117.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 100.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 136.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 126.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 141.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 105.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 115.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 97.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 135.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 125.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 144.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 109.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 120.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 102.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 138.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 105.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:33 120.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:40 108.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:34 149.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:29 136.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:30 122.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:37 110.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 94.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 109.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:36 124.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:38 112.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:33 152.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:29 140.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:36 126.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:35 114.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 98.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 101.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 89.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 104.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 92.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 102.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 90.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 105.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 93.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 113.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 100.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 116.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 103.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 141.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 125.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 136.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 118.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 143.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 128.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 140.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 121.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 122.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 122.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 131.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 116.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 125.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 125.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 135.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 123.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 111.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 122.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 106.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 133.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 90.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 100.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 126.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 122.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 114.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 110.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 139.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 125.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 93.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 115.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 103.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 134.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 129.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 137.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 134.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 115.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 124.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 109.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 119.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 119.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 129.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 103.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 113.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 116.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 105.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 99.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 99.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 97.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 84.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 106.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 94.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 119.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 115.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 107.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 103.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 104.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 132.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 100.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 87.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 131.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 114.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 126.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 109.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 107.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 134.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 131.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 112.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 121.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 97.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 106.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 126.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 100.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 110.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 113.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 109.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 102.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 96.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 112.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 96.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 93.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 81.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 102.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 90.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 116.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 104.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 116.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 101.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 129.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 97.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 84.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 95.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 133.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 128.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 110.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 136.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 121.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 133.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 114.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 114.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 123.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 99.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 119.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 118.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 128.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 102.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 116.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 110.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 103.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 98.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 114.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 97.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 125.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 112.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 95.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 83.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 104.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 92.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 119.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 106.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 118.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 131.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 118.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 98.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 86.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 126.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 112.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 129.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 115.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 118.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 105.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 121.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 108.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 115.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 103.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 118.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 105.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 116.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 104.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 119.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 107.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 148.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 129.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 140.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 124.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 122.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 151.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 132.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 145.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 125.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 130.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 126.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 135.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 120.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 133.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 130.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 140.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 124.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 139.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 127.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 123.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 113.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 127.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 110.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 140.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 126.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 108.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 96.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 120.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 104.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 140.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 129.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 126.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 131.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 145.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 130.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 110.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 99.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 124.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 141.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 122.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 134.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 115.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 145.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 126.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 139.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 119.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 123.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 119.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 129.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 105.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 127.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 124.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 134.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 117.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 131.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 116.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 107.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 103.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 133.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 118.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 101.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 90.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 97.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 132.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 123.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 119.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 124.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 138.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 93.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 138.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 131.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 142.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 123.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 136.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 116.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 120.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 116.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 126.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 102.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 111.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 124.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 131.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 114.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 128.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 116.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 113.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 116.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 130.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 115.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 98.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 87.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 94.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 129.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 116.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 121.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 135.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 121.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 90.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 114.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 98.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 141.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 121.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 133.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 117.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 144.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 126.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 139.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 120.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 118.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 123.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 119.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 129.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 105.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 127.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 123.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 134.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 131.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 118.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 115.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 118.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 130.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 117.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 101.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 89.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 108.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 132.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 122.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 118.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 110.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 123.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 136.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 123.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 104.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 92.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 102.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 110.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 97.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 113.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 100.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 111.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 98.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 114.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 101.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 131.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 132.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 140.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 135.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 126.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 120.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 143.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 136.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 129.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 122.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 120.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 107.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 119.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 103.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 123.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 106.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 128.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 123.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 115.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 110.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 109.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 95.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 131.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 118.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 111.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 99.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 118.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 102.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 124.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 118.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 111.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 105.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 104.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 90.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 126.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 122.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 113.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 109.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 107.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 94.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 141.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 127.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 144.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 130.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 129.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 116.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 132.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 119.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 125.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 111.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 128.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 114.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 139.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 121.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 140.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 143.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 129.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 125.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 144.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 130.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 125.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 112.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 127.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 109.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 130.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 131.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 100.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 134.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 116.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 122.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 104.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 126.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 109.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 96.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 130.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 116.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 112.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 171.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 162.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 172.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 163.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 167.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 180.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 160.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 168.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 183.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 161.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 167.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 150.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 184.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 169.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 155.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 143.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 170.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 151.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 188.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 173.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 156.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 144.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 155.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 146.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 160.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 151.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 151.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 165.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 144.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 156.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 172.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 149.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 152.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 133.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 169.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 154.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 140.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 126.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 157.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 139.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 175.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 160.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 145.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 132.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 148.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 139.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 154.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 145.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 144.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 158.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 137.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 150.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 165.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 143.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 146.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 163.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 147.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 133.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 120.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 151.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 134.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 170.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 154.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 139.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 126.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 175.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 166.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 178.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 169.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 171.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 184.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 164.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 175.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 188.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 167.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 171.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 154.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 191.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 174.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 164.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 147.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 174.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 154.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 195.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 177.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 166.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 147.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 159.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 150.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 165.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 155.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 155.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 169.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 148.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 161.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 176.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 153.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 156.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 137.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 176.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 158.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 149.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 130.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 161.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 143.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 182.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 165.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 153.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 136.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 153.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 144.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 159.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 149.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 149.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 163.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 142.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 155.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 170.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 147.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 150.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 131.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 169.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 151.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 142.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 123.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 156.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 137.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 176.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 158.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 147.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 130.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 94.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 82.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 98.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 86.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 94.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:35 104.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:30 92.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:38 132.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:34 121.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:32 106.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:35 94.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 82.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 98.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:36 107.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:39 96.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:30 136.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:40 124.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:35 110.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:37 98.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 86.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 105.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 93.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 109.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 97.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 103.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 100.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 123.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 110.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 104.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 108.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 102.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 112.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 100.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 99.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 114.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 103.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 106.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 94.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 90.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 111.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 98.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 115.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 102.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 98.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 83.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 86.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 111.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 99.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 95.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 115.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 102.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 120.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 101.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 86.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 89.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 114.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 99.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 95.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 120.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 103.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 114.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 108.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 99.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 105.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 97.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 94.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 111.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 100.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 99.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 103.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 91.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 86.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 108.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 95.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 98.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 95.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 80.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 93.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 81.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 108.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 96.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 102.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 90.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 112.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 99.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 103.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 99.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 83.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 97.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 85.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 113.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 97.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 103.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 93.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 106.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 104.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 96.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 106.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 96.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 92.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 100.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 99.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 96.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 90.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 96.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 84.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 106.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 94.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 96.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 93.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 78.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 91.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 79.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 100.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 88.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 110.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 98.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 113.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 101.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 97.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 82.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 95.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 83.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 100.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 107.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 122.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 105.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 115.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 109.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 101.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 99.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 99.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 95.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 112.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 113.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 102.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 104.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 92.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 86.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 108.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 95.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 110.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 98.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 96.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 81.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 93.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 81.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 109.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 97.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 102.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 91.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 112.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 99.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 116.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 99.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 84.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 97.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 86.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 122.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 102.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 112.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 98.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 127.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 101.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 112.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 97.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 105.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 93.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 99.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 87.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 109.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 97.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 96.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 84.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 98.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 98.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 107.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 94.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 109.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 97.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 108.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 93.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 90.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 95.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 83.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 105.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 93.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 91.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 79.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 107.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 95.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 117.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 95.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 105.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 92.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 123.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 94.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 106.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 91.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 113.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 100.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 89.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 93.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 81.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 103.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 91.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 89.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 77.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 94.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 98.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 108.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 95.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 126.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 111.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 97.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 109.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 94.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 116.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 91.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 95.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 83.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 105.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 93.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 91.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 79.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 96.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 109.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 98.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 113.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 102.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 106.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 95.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 110.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 99.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 104.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 94.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 108.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 98.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 106.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 95.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 110.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 99.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 119.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 104.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 130.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 112.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 124.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 108.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 115.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 107.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 117.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 106.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 103.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 111.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 122.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 109.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 113.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 102.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 94.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 125.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 112.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 120.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 107.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 103.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 88.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 89.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 118.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 110.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 126.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 114.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 112.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 107.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 91.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 105.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 93.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 122.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 111.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 128.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 120.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 114.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 104.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 112.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 107.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 118.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 111.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 99.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 102.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 90.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 122.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 109.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 100.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 85.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 96.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 85.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 94.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 123.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 111.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 120.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 104.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 89.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 89.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 120.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 113.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 97.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 126.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 117.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 101.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 111.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 102.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 97.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 117.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 105.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 100.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 97.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 88.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 120.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 108.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 101.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 99.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 84.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 94.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 83.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 114.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 104.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 92.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 122.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 109.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 118.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 102.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 87.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 99.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 86.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 124.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 106.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 117.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 102.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 130.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 110.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 122.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 115.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 106.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 120.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 120.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 108.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 105.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 112.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 100.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 103.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 122.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 110.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 102.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 90.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 86.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 105.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 124.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 111.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 122.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 109.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 105.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 92.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 92.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 130.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 107.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 119.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 102.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 128.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 119.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 127.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 103.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 115.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 98.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 125.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 100.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 112.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 96.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 124.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 114.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 130.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 104.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 117.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 100.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 129.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:40 104.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:31 92.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:32 133.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:38 121.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:37 106.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:37 94.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:38 108.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:32 96.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:30 136.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:36 124.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:40 110.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:30 98.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 100.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 88.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 104.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 92.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 100.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:39 107.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:34 94.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:30 135.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:32 123.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:33 109.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:29 96.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 88.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 104.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:38 111.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:40 98.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:37 139.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:34 126.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:32 113.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:30 100.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 92.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 99.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 114.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 103.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 131.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 108.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 118.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 115.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 101.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 134.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 123.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 117.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 105.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 123.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 107.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 115.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 122.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 111.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 120.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 107.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 118.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 106.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 93.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 117.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 105.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 107.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 103.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 91.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 100.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 87.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 120.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 113.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 98.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 121.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 109.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 94.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 91.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 128.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 104.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 112.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 97.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 132.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 119.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 114.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 120.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 111.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 102.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 96.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 105.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 100.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 89.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 114.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 102.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 116.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 88.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 95.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 83.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 93.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 117.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 106.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 108.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 91.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 87.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 127.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 110.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 95.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 130.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 117.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 113.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 99.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 118.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 100.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 109.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 101.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 94.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 104.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 114.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 102.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 87.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 112.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 101.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 101.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 99.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 87.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 93.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 81.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 116.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 105.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 106.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 91.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 116.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 104.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 119.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 89.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 97.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 85.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 130.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 98.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 133.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 109.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 120.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 116.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 102.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 122.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 97.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 107.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 101.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 116.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 103.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 89.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 113.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 101.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 103.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 101.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 89.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 95.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 83.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 118.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 94.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 117.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 105.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 121.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 92.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 99.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 87.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 135.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 107.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 116.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 100.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 138.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 127.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 106.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 114.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 99.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 116.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 105.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 105.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 90.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 114.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 102.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 97.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 85.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 133.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 102.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 112.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 96.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 135.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 101.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 110.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 95.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 122.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 114.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 102.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 101.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 86.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 110.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 98.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 93.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 81.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 116.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 105.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 131.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 100.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 109.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 93.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 134.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 123.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 99.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 108.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 93.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 121.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 101.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 98.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 84.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 107.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 95.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 91.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 79.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 104.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 135.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 103.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 113.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 96.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 137.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 126.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 102.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 111.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 95.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 124.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 101.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 86.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 110.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 98.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 93.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 81.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 105.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 115.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 105.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 119.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 109.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 112.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 102.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 116.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 106.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 111.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 101.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 115.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 105.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 112.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 101.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 116.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 105.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 138.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 113.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 123.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 121.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 142.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 128.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 124.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 130.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 120.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 129.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 125.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 114.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 109.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 125.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 97.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 131.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 118.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 125.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 109.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 97.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 127.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 117.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 132.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 120.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 130.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 99.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 94.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 136.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 119.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 118.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 139.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 113.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 124.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 121.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 105.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 127.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 107.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 116.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 108.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 100.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 126.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 111.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 121.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 111.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 122.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 110.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 93.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 128.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 116.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 121.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 108.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 106.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 87.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 125.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 113.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 97.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 129.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 117.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 113.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 109.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 97.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 103.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 90.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 134.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 106.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 117.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 99.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 137.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 122.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 126.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 114.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 107.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 98.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 125.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 109.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 119.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 109.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 102.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 121.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 109.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 126.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 114.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 104.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 92.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 96.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 84.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 123.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 111.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 111.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 95.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 127.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 116.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 123.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 95.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 101.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 88.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 138.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 110.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 121.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 121.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 141.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 127.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 123.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 129.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 118.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 102.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 128.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 114.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 124.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 113.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 124.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 111.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 110.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 94.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 128.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 116.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 122.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 109.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 107.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 95.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 100.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 88.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 126.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 115.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 101.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 130.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 118.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 127.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 114.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 110.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 98.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 106.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 94.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 145.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 112.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 122.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 104.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 140.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 130.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 142.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 107.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 118.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 100.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 137.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 127.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 141.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 105.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 115.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 97.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 136.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 126.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 144.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 109.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 120.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 102.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 139.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 129.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 100.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:39 107.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:30 94.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:39 135.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:35 123.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:29 109.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:37 96.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 88.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 104.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:38 111.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:35 98.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:38 139.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:31 126.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:40 113.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:29 100.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 92.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 96.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 85.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 99.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 87.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 97.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 85.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 100.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 88.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 108.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 95.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 98.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 127.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 125.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 136.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 118.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 130.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 128.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 140.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 120.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 121.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 115.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 122.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 131.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 107.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 116.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 125.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 135.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 109.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 118.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 106.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 101.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 125.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 111.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 139.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 126.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 105.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 92.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 113.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 121.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 109.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 104.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 129.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 143.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 130.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 107.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 95.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 116.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 129.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 134.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 114.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 108.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 124.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 100.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 109.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 119.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 129.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 103.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 113.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 111.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 100.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 95.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 118.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 132.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 119.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 98.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 86.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 94.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 114.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 110.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 103.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 122.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 137.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 101.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 89.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 110.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 126.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 121.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 119.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 131.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 111.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 105.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 112.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 97.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 106.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 109.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 116.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 100.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 110.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 108.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 104.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 97.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 91.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 115.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 100.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 129.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 115.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 95.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 83.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 103.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 91.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 111.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 100.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 95.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 119.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 105.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 134.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 98.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 86.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 95.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 128.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 110.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 110.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 123.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 121.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 133.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 114.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 107.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 114.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 123.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 99.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 111.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 118.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 129.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 111.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 99.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 93.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 116.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 130.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 115.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 97.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 85.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 104.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 92.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 114.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 109.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 102.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 97.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 105.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 136.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 121.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 100.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 88.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 108.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 97.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 121.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 107.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 124.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 110.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 113.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 101.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 116.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 104.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 110.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 113.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 101.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 111.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 99.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 114.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 102.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 135.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 129.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 140.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 123.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 122.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 138.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 132.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 145.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 125.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 125.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 123.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 135.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 112.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 120.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 125.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 130.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 140.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 124.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 134.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 122.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 118.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 129.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 115.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 144.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 128.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 110.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 104.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 135.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 121.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 112.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 133.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 118.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 147.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 134.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 124.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 128.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 122.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 134.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 132.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 139.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 119.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 120.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 129.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 105.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 114.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 119.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 134.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 108.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 117.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 126.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 112.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 122.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 105.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 137.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 122.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 103.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 91.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 97.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 127.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 118.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 114.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 106.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 126.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 141.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 102.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 125.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 119.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 131.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 112.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 129.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 123.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 136.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 116.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 113.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 126.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 102.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 111.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 121.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 131.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 105.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 114.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 123.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 111.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 109.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 99.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 102.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 133.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 100.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 88.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 110.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 94.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 124.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 115.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 111.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 103.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 123.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 107.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 138.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 124.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 103.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 91.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 99.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 128.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 121.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 133.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 114.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 132.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 126.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 139.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 119.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 118.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 115.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 119.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 129.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 105.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 119.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 123.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 134.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 108.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 117.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 126.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 111.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 101.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 121.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 135.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 103.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 90.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 109.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 126.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 117.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 114.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 105.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 125.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 111.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 140.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 126.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 106.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 113.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 102.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 105.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 93.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 108.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 95.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 106.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 93.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 109.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 96.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 124.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 115.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 125.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 116.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 135.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 130.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 122.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 122.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 138.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 131.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 124.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 122.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 109.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 103.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 115.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 106.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 123.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 118.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 111.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 110.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 97.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 126.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 122.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 113.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 109.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 113.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 100.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 107.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 111.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 118.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 113.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 106.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 101.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 92.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 121.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 109.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 104.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 95.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 137.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 123.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 139.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 125.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 124.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 112.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 127.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 114.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 120.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 107.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 122.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 110.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 131.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 133.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 121.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 138.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 125.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 127.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 139.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 114.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 123.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 126.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 113.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 115.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 102.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 129.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 118.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 115.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 119.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 121.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 108.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 110.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 97.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 125.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 112.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 171.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 162.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 172.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 163.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 167.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 180.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 160.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 168.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 183.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 161.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 171.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 156.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 191.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 177.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 156.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 144.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 174.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 158.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 196.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 180.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 157.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 144.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 155.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 146.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 160.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 151.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 151.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 165.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 144.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 156.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 172.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 149.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 156.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 140.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 176.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 161.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 140.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 127.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 161.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 145.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 182.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 167.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 145.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 133.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 148.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 139.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 154.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 145.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 144.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 158.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 137.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 150.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 165.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 143.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 150.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 133.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 170.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 154.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 134.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 155.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 139.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 176.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 161.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 139.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 127.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 175.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 166.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 179.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 169.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 171.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 184.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 164.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 175.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 188.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 167.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 175.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 160.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 196.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 180.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 164.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 147.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 178.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 162.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 200.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 184.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 166.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 147.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 159.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 150.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 165.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 155.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 156.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 169.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 148.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 161.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 176.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 153.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 160.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 144.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 181.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 163.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 149.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 130.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 165.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 149.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 187.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 170.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 154.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 136.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 153.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 144.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 159.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 149.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 149.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 163.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 142.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 155.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 170.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 147.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 153.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 138.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 174.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 156.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 143.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 160.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 143.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 180.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 164.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 148.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 130.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 95.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 83.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 99.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 87.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 95.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:33 111.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:33 87.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:32 138.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:39 116.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:37 113.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:34 89.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 83.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 99.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:35 111.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:30 91.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:34 140.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:36 120.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:39 113.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:32 93.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 87.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 106.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 94.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 110.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 98.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 117.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 100.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 123.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 118.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 100.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 99.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 114.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 116.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 103.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 103.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 95.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 103.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 90.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 112.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 99.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 99.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 84.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 86.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 112.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 100.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 95.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 116.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 103.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 121.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 108.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 87.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 90.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 114.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 110.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 95.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 120.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 108.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 99.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 105.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 98.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 108.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 97.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 94.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 111.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 112.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 100.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 99.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 105.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 92.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 98.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 86.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 109.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 96.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 99.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 96.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 81.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 93.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 81.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 109.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 97.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 103.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 91.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 113.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 100.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 117.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 104.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 99.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 84.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 97.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 85.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 113.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 97.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 107.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 103.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 93.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 119.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 112.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 97.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 103.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 96.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 96.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 92.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 109.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 100.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 110.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 99.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 96.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 103.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 91.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 96.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 84.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 107.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 95.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 110.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 97.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 79.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 91.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 79.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 108.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 95.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 100.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 88.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 111.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 99.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 114.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 102.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 98.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 83.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 95.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 83.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 116.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 110.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 96.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 122.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 101.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 107.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 99.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 109.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 99.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 95.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 112.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 93.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 86.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 109.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 96.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 112.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 99.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 96.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 81.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 93.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 82.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 110.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 98.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 112.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 100.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 117.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 100.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 85.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 97.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 86.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 122.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 102.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 112.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 98.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 127.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 101.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 112.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 97.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 117.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 94.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 100.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 88.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 110.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 98.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 96.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 84.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 111.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 98.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 98.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 107.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 94.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 124.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 109.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 97.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 108.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 93.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 103.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 91.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 95.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 83.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 106.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 94.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 91.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 80.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 96.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 95.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 105.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 92.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 123.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 94.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 106.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 91.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 113.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 102.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 89.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 93.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 81.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 103.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 92.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 89.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 77.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 106.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 121.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 98.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 108.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 95.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 126.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 111.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 97.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 109.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 94.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 116.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 104.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 92.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 95.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 83.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 106.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 94.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 91.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 80.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 96.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 110.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 99.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 114.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 103.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 107.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 96.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 111.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 100.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 105.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 95.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 109.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 99.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 106.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 96.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 110.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 100.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 124.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 108.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 119.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 114.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 104.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 131.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 112.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 124.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 117.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 107.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 117.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 107.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 121.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 111.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 122.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 109.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 107.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 102.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 107.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 94.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 126.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 113.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 104.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 89.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 89.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 119.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 107.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 127.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 114.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 125.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 108.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 92.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 106.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 93.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 122.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 115.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 100.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 128.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 108.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 120.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 114.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 104.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 112.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 113.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 104.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 99.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 118.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 107.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 118.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 103.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 112.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 99.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 102.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 90.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 123.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 110.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 104.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 101.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 86.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 97.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 85.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 104.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 106.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 94.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 124.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 112.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 105.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 89.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 101.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 89.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 120.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 102.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 113.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 110.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 97.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 126.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 112.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 101.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 101.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 111.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 102.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 105.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 105.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 110.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 98.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 100.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 88.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 121.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 109.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 102.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 100.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 84.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 94.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 83.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 115.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 103.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 92.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 123.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 110.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 119.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 107.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 103.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 88.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 99.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 87.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 124.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 106.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 117.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 114.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 102.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 130.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 122.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 115.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 105.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 106.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 101.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 120.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 110.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 120.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 108.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 105.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 113.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 100.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 91.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 123.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 111.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 118.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 103.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 90.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 98.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 86.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 118.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 106.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 98.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 125.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 112.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 123.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 110.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 106.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 93.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 92.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 130.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 107.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 119.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 102.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 130.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 119.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 127.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 103.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 115.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 98.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 127.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 116.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 126.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 100.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 113.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 96.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 125.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 130.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 105.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 117.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 100.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 130.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 118.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:31 111.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:33 87.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:37 138.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:36 116.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:39 113.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:39 89.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:36 111.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:40 91.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:35 140.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:31 120.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:37 114.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:35 93.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 106.37 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 94.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 111.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 100.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 105.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 94.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 110.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 100.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 116.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 105.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 122.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 111.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 139.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.89 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 133.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 126.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 115.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 149.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 127.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 138.54 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 129.43 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 119.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 131.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 130.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 116.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 136.93 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 125.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 135.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 119.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 118.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 127.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 113.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 124.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 128.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 116.10 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 137.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 121.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 101.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 115.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 101.67 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 136.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 123.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 127.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 112.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 133.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 121.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 140.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 126.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 118.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 104.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 105.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 134.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 125.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 121.17 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.10 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 144.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 120.22 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 131.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 124.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 112.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 113.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 111.08 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 132.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 127.53 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 114.43 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 111.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 121.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 108.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 116.60 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 98.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 122.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 110.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 127.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 113.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 110.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 96.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 92.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 130.22 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.78 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 120.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.08 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 127.57 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 116.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 133.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 99.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 130.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 110.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 120.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 102.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 141.40 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.47 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 125.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 120.63 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 123.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 108.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 117.35 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 107.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 101.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 128.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 112.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 122.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 110.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 106.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 105.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 111.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 94.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 119.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 107.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 122.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 106.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 92.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 100.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 87.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 98.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 124.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 112.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 127.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 113.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 96.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 132.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 112.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 122.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 119.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.82 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 143.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 116.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 128.15 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 122.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 109.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 125.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.53 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 119.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 109.89 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 104.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 130.74 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 115.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 125.22 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 113.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.25 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 95.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 119.82 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 107.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 123.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 109.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 101.86 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 88.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 128.28 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 116.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.43 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 100.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 124.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 113.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 128.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 111.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 97.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 105.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 93.17 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 142.82 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 121.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 130.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 114.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 151.22 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 134.51 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 120.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 128.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 113.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 138.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 126.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 113.54 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 120.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 105.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 130.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 114.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 113.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 99.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 133.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 121.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 137.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 114.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 122.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 106.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 146.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 129.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 112.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 121.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 106.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 133.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 120.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 108.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 112.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 97.28 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 121.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 106.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 104.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 91.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 116.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 134.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 108.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 117.67 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 101.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 142.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 126.02 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 107.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 116.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 100.87 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 130.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 116.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 104.74 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 107.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 92.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 116.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 101.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 98.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 87.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 123.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 135.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 110.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 119.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 103.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 144.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 127.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 109.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 118.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 102.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 132.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.11 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 108.08 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 93.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 116.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 102.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 99.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 88.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 125.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 114.65 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 127.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 116.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 132.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 121.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 121.99 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 110.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 126.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 115.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 118.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 107.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 123.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 112.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 119.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 108.00 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 124.11 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 113.54 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 147.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 127.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 138.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 132.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 119.99 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 155.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 131.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 143.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 135.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 123.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 139.00 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 134.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 122.37 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 119.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 142.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 129.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 139.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 125.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 122.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 134.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 121.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 128.35 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 110.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 139.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 127.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 140.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 125.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 120.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 107.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 104.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 141.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 129.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 132.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 116.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 144.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 132.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 144.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 130.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 124.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 121.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 141.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 120.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 130.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 127.87 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 112.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 150.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 135.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 130.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 116.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 133.74 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 118.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 127.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 117.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 111.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 137.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 122.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 132.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 116.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 128.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 115.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 120.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 102.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 133.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 122.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 132.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 115.67 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 109.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 95.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 135.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 123.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 125.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.78 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 138.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 127.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 137.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 123.57 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 119.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 105.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 114.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 99.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 138.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 114.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 125.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 124.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 147.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 119.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 130.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 126.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 111.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 130.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 113.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 122.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.00 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 134.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 117.53 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 127.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.89 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 110.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 125.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 114.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 130.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 118.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 126.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 112.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 111.76 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 91.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 132.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 120.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 119.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 102.25 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 135.60 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 123.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 131.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 118.02 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 115.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 101.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 108.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 95.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 140.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 128.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 126.47 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 110.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 149.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 122.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 134.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 129.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 114.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 133.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 116.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 125.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 116.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 109.65 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 136.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 121.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 131.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 119.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 113.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 127.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 117.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 100.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 131.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 120.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 128.93 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 114.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 100.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 106.67 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 93.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 134.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 122.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 122.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 136.86 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 125.40 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 134.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 120.53 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 103.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 99.76 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 151.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 126.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 136.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 118.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 147.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 133.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 132.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 146.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 118.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 128.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 110.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 141.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 127.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 143.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 113.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 123.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 105.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 137.86 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 123.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 145.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 116.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 127.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 108.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 139.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 126.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 125.02 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 106.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 94.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 111.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 100.29 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 102.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 87.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 105.13 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 91.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 102.08 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 90.97 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 105.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 93.93 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 112.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 98.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 115.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 101.53 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 147.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 151.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 163.96 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 138.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 144.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 153.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 152.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 165.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 140.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 144.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 134.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 148.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 159.04 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 127.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 142.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 140.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 149.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 161.08 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 129.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 143.53 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 142.92 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 134.04 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 123.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 117.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 145.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 130.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 162.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 145.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 128.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 135.39 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 145.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 136.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 126.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 123.97 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 149.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 134.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 164.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 150.43 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 138.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 124.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 132.65 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 136.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 147.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 123.44 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 129.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 138.25 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 139.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 152.00 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 132.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 133.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 142.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 112.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 127.46 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 125.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 136.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 147.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 114.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 130.42 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 123.72 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 107.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 101.92 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 133.60 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 117.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 147.46 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 132.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 111.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 95.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 121.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 126.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 123.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 110.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 108.43 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 137.37 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 120.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 151.74 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 136.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 124.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 110.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 125.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 128.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 140.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 116.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 121.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 132.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 132.78 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 145.93 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 119.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 125.56 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 126.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 136.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 105.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 120.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 119.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 130.36 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 141.15 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 124.01 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 115.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 110.58 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 101.60 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 95.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 126.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 140.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 125.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 103.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 89.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 100.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 118.69 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 117.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 104.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 102.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 130.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 114.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 145.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 130.72 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.67 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 92.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 117.91 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 128.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 141.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 122.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 132.93 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 133.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 146.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 120.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 113.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 126.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 136.39 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 120.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 119.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 131.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 141.97 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 109.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 124.87 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 116.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 110.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 102.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 96.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 125.96 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 108.57 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 139.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 103.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 89.67 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 113.67 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 100.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 119.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 117.08 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 105.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 103.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 130.58 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 113.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 145.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 129.69 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 93.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.13 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 105.14 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 143.23 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 128.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 146.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 131.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 124.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 112.25 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 127.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 115.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 116.64 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 105.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 119.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 108.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 116.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 105.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 119.47 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 108.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 155.50 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 154.78 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 168.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 143.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 147.57 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 157.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 156.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 170.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 145.60 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 148.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 142.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 152.40 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 163.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 133.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 146.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 144.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 153.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 165.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 134.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 147.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 154.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 141.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 134.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 124.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 149.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 134.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 167.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 150.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 133.36 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 117.08 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 140.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 125.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 157.03 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 141.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 137.40 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 128.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 153.44 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 138.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 170.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 154.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 131.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 143.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 127.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 140.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 140.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 152.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 128.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 133.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 142.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 144.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 156.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 130.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 136.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 127.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 137.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 147.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 117.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 131.81 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 129.65 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 141.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 151.78 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 120.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 134.99 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 135.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 125.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 118.83 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 109.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 137.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 119.00 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 152.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 136.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 116.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 100.37 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 128.53 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 110.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 138.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 128.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 121.64 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.94 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 141.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 122.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 156.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 140.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 118.28 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 103.23 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 131.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 113.67 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 133.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 133.33 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 145.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 121.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 125.99 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 136.82 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 137.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 150.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 124.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 129.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 120.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 130.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 140.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 111.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.56 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 123.61 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 135.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 145.74 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 128.33 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 127.61 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 117.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 112.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 103.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 129.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 112.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 144.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 129.50 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 108.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 94.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 120.83 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 104.04 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 130.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 121.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 115.82 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 134.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 150.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 134.44 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 97.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 134.28 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 133.92 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 146.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 122.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 126.72 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 138.00 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 138.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 151.96 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 125.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 130.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 121.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 131.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 142.25 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 125.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 124.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 135.91 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 147.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.23 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 129.43 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 128.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 118.25 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 114.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 130.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 114.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 145.01 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 129.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 109.61 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 95.47 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 118.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.49 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 130.99 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 122.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 117.15 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 108.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 134.86 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 119.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 149.93 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 134.74 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 112.72 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 98.78 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 122.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 110.55 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 110.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 98.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 112.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 101.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 109.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 101.81 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 112.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 104.65 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 155.09 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 156.12 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 156.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 154.24 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 160.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 151.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 151.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 146.24 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 155.06 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 147.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 163.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 154.93 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 154.13 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 145.59 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 152.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 141.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 130.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 127.34 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 135.17 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 129.68 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 137.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 132.70 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 124.87 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 131.77 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 118.77 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 140.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 136.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 127.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 128.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 133.18 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 120.03 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 114.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 124.18 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 124.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 119.23 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 113.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.00 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 118.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 105.15 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 127.20 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 126.11 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 116.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 121.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 108.10 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 159.61 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 150.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 162.47 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 153.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 139.51 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 129.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 142.37 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 132.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 125.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 116.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 128.19 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 119.91 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 162.64 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 161.54 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 160.71 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 159.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 160.77 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 153.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 160.19 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 152.28 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 157.61 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 150.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 157.82 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 145.97 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 137.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 132.64 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 139.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 134.97 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 140.88 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 126.89 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 136.91 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 123.44 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 141.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 132.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 138.32 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 124.76 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 124.97 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 119.81 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.69 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 122.31 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 126.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 115.34 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 123.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 110.34 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 130.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 119.35 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 126.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 113.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 231.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 222.70 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 228.55 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 219.04 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 227.26 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 241.71 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 219.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 224.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 240.36 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 216.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 202.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 192.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 225.25 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 214.06 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 190.93 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 184.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 198.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 194.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 225.10 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 216.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 186.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 183.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 193.23 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 183.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 199.09 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 189.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 188.81 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 205.06 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 181.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 194.90 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 210.04 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 187.11 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 177.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 167.43 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 202.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 189.52 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 164.48 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 155.68 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 180.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 167.90 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 205.57 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 190.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 166.37 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 158.20 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 178.15 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 168.89 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 183.59 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 174.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 173.72 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 189.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 166.31 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 179.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 195.65 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 171.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 163.20 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 150.70 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 186.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 172.43 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 149.56 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 138.91 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 169.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 156.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 193.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 179.00 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 155.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 144.87 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 236.36 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 226.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 232.96 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 223.05 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 231.78 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 246.46 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 224.12 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 229.13 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 244.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 221.08 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 206.82 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 196.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 230.72 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 221.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 195.57 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 188.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 203.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 198.62 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 235.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 220.74 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 191.29 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 186.59 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 197.63 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 188.12 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 203.62 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 193.72 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 193.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 209.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 185.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 199.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 214.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 191.25 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 182.23 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 171.68 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 206.62 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 192.90 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 170.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 159.25 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 184.16 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 172.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 212.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 194.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 172.51 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 161.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 182.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 173.03 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 187.99 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 178.21 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 178.24 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 194.05 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 170.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 183.71 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 200.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 175.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 167.18 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 154.83 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 190.55 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 176.13 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 155.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 142.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 173.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 160.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 201.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 182.48 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 161.52 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 148.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 102.50 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 90.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 109.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 98.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 101.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 90.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 109.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 97.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 113.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 101.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 120.36 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 108.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 123.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 107.74 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 117.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 103.86 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 133.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.42 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 122.33 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 114.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 107.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 106.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 115.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 123.81 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 110.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 120.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 113.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 101.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 94.78 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 118.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 106.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 119.74 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 107.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.17 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 87.50 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 101.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 89.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 122.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 110.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 111.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 99.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 126.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 114.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 125.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 112.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 106.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 91.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 105.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 93.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 120.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 103.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 113.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 108.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 99.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 130.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 107.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 118.87 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 112.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 103.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 111.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 111.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 101.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 121.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 116.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 104.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 102.61 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 110.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 90.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 115.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 103.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 99.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 84.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 97.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 85.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 119.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 107.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 106.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 94.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 123.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 111.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 121.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 108.72 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 103.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 88.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 89.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 118.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 101.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 111.24 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 107.36 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.35 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 129.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 105.67 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 116.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 110.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 101.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 109.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 100.39 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 109.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 99.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 96.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 119.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 104.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 114.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.78 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 100.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 108.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 96.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 88.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 114.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 102.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 113.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 101.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 98.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 83.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 95.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 83.37 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 117.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 106.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 104.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 92.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 121.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 110.03 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 106.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.01 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 87.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 99.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 87.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 122.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 114.21 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 111.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 100.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 132.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 109.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 119.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 113.87 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 113.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.24 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 103.45 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 99.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 122.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 118.18 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 106.35 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 103.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 111.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 98.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 102.53 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 90.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 115.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 103.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 115.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 103.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 100.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 85.39 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 97.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 85.78 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 107.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 107.10 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 95.29 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 123.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 111.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 109.21 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 89.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 101.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 89.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 126.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 105.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 115.53 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 102.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 134.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 117.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 104.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 114.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 100.94 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 125.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 112.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 100.28 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 103.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 92.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 114.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 100.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 99.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 87.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 107.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 123.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 102.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 112.08 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 98.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 132.10 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 114.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 100.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 110.44 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 97.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 122.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 97.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 99.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 87.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 110.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 96.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 95.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 83.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 116.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 104.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 121.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 100.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 109.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 96.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 130.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 98.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 108.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 95.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 120.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 107.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 96.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 97.21 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 85.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 108.33 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 93.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 93.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 81.35 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 114.86 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 103.45 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 125.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 102.72 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 112.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 99.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 133.44 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.10 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 101.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 111.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 97.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 123.86 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 109.83 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 97.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 99.39 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 87.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 110.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 96.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 95.21 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 83.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 116.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 106.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 125.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 114.21 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 114.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 103.83 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 122.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 111.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 112.81 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 102.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 121.08 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 110.03 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 114.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 103.37 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 122.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 111.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 130.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 112.44 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 123.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 118.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 139.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 116.01 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 127.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 121.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 111.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 121.45 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 110.58 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 107.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 129.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 115.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 125.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 112.98 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 110.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 120.61 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 108.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.87 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 130.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 117.83 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 124.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 112.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 107.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 92.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 105.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 93.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 127.76 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 115.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 114.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 102.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 137.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 124.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 129.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 117.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 111.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 96.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 109.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 96.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 127.85 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 118.90 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 115.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.85 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 136.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 112.42 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 124.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 118.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 118.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 107.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 117.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 107.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 102.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 126.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 122.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 110.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 106.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 117.74 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 105.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.37 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 94.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 127.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 115.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 120.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 108.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 105.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 89.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 100.94 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 89.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 125.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 112.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 110.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 98.67 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 134.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 122.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 125.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 108.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 93.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 105.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 92.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 126.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 105.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 116.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 134.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 110.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 121.87 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 116.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 105.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 117.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 104.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 114.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 100.72 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 125.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 109.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 120.08 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 109.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 104.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 103.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 104.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 92.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 126.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 114.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 118.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 106.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 103.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 88.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 98.51 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 86.82 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 123.61 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 111.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.33 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 96.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 132.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 120.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 123.58 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 111.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 107.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 92.39 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 90.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 130.17 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 121.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 117.98 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 105.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 138.36 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 114.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 126.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 120.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 110.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 110.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 104.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 128.53 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 114.03 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 124.82 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 112.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 109.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 119.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 107.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 95.78 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 128.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 116.28 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 122.19 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 109.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 106.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 94.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 101.94 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 90.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 114.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 112.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 102.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 134.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 122.76 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 127.85 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 114.98 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 110.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 98.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 108.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 96.39 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 134.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 111.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 123.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 106.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 133.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 123.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 122.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 131.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 106.85 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 118.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 102.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 130.42 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 120.29 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.90 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 129.69 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 104.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 116.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 99.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 128.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 118.76 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 118.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 133.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 108.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 120.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 104.01 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 133.86 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 121.63 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 120.52 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 106.77 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 90.05 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 113.19 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 97.20 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 107.05 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 89.79 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 114.39 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 96.90 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 117.44 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 100.97 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 123.89 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 108.02 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 128.33 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 111.70 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 134.88 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 118.73 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 122.66 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 106.33 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 129.08 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 113.36 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 119.42 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 102.82 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 125.84 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 109.85 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 128.97 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 111.31 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 136.27 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 118.47 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 122.34 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 105.65 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 129.77 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 112.68 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 119.22 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 102.70 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 126.66 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 109.73 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 139.87 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 123.45 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 146.31 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 129.34 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 134.20 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 118.06 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 140.63 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 124.09 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 130.84 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 114.43 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 137.27 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 120.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 91.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 79.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 87.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 85.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 74.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 100.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 89.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 96.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 94.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 84.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 85.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 78.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 84.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 80.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 74.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 94.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 87.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 94.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 89.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 84.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 78.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 70.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 74.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 82.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 72.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 73.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 64.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 67.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 59.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 84.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 75.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 80.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 69.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 88.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 79.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 79.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 71.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 73.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 66.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 89.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 77.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 84.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 82.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 72.67 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 98.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 87.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 94.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 92.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 82.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 83.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 75.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 81.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 78.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 72.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 92.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 85.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 92.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 87.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 81.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 76.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 67.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 71.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 60.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 79.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 70.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 71.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 61.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 64.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 56.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 81.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 73.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 78.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 67.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 86.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 77.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 77.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 68.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 71.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 63.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 87.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 76.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 83.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 81.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 71.42 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 97.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 85.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 93.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 91.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 81.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 82.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 74.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 80.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 76.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 70.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 91.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 84.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 90.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 86.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 80.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 74.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 66.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 70.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 59.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 78.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 69.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 69.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 60.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 63.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 55.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 80.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 72.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 76.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 66.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 85.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 76.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 76.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 67.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 70.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 62.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 91.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 79.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 86.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 84.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 73.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 100.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 89.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 96.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 82.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 85.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 77.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 83.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 80.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 72.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 94.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 87.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 94.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 89.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 82.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 76.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 68.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 72.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 61.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 80.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 71.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 71.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 65.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 57.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 82.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 74.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 78.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 68.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 86.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 78.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 78.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 69.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 72.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 64.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 78.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 85.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 74.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 103.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 88.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 95.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 84.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 88.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 77.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 83.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 73.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 97.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 87.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 93.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 83.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 76.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 71.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 69.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 64.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 77.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 72.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 66.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 61.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 82.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 80.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 75.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 74.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 83.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 81.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 72.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 71.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 92.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 76.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 83.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 72.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 101.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 86.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 93.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 81.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 86.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 75.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 81.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 71.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 95.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 85.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 91.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 81.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 74.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 69.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 66.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 61.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 74.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 69.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 63.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 59.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 80.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 78.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 73.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 71.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 81.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 79.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 70.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 68.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 91.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 75.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 82.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 70.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 100.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 84.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 92.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 80.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 85.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 73.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 79.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 70.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 93.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 84.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 89.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 79.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 72.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 68.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 65.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 60.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 73.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 68.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 62.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 58.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 79.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 77.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 72.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 70.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 80.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 78.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 69.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 67.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 94.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 78.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 85.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 73.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 103.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 87.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 94.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 83.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 88.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 77.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 82.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 71.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 97.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 86.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 92.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 81.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 74.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 70.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 67.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 75.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 70.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 64.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 60.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 80.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 79.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 74.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 72.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 82.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 80.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 71.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 69.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 78.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 73.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 88.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 83.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 76.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 73.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 86.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 82.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 68.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 64.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 65.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 61.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 75.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 74.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 72.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 71.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 76.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 71.53 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 85.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 81.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 74.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 70.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 84.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 80.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 65.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 61.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 63.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 59.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 72.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 71.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 69.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 68.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 74.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 70.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 84.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 80.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 73.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 69.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 83.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 79.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 64.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 60.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 61.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 58.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 71.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 70.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 68.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 67.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 78.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 73.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 88.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 83.42 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 76.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 71.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 86.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 81.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 66.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 63.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 60.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 73.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 72.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 70.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 69.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 98.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 83.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 81.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 91.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 90.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 78.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 76.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 107.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 92.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 91.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 99.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 87.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 86.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 92.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 81.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 79.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 88.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 85.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 77.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 76.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 91.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 89.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 98.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 87.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 85.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 85.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 76.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 77.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 73.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 66.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 62.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 86.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 77.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 78.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 69.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 72.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 69.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 62.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 59.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 90.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 82.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 83.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 80.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 73.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 69.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 92.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 84.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 84.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 76.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 78.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 75.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 68.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 66.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 96.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 80.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 79.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 89.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 88.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 75.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 74.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 105.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 90.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 89.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 98.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 97.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 85.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 84.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 90.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 79.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 77.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 86.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 83.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 75.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 73.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 89.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 87.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 96.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 92.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 85.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 83.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 83.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 74.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 74.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 71.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 63.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 60.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 84.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 74.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 76.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 66.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 69.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 67.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 59.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 57.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 88.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 80.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 81.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 77.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 71.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 67.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 90.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 81.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 82.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 73.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 76.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 73.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 66.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 64.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 94.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 79.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 77.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 87.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 86.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 74.67 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 73.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 89.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 87.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 97.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 96.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 84.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 82.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 89.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 78.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 76.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 84.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 81.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 73.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 72.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 98.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 88.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 86.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 94.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 91.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 83.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 82.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 81.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 73.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 73.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 69.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 59.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 83.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 73.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 74.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 65.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 68.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 65.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 58.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 56.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 87.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 79.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 80.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 76.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 69.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 66.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 89.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 80.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 80.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 72.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 75.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 72.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 65.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 63.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 98.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 83.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 81.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 91.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 90.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 76.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 77.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 108.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 93.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 91.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 100.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 99.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 86.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 86.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 92.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 82.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 80.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 88.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 85.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 75.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 74.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 102.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 90.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 95.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 85.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 84.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 84.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 75.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 76.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 70.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 65.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 61.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 85.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 76.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 77.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 68.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 69.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 67.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 61.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 59.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 90.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 82.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 82.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 77.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 72.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 68.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 91.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 83.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 83.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 75.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 75.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 73.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 68.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 66.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 115.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 92.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 100.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 86.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 104.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 96.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 113.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 89.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 97.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 84.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 94.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 111.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 88.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 96.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 83.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 101.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 93.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 116.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 92.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 100.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 85.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 103.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 95.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 98.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 95.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 119.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 127.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 126.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 97.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 103.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 100.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 94.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 125.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 117.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 124.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 120.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 114.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 93.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 81.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 104.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 92.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 94.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 84.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 85.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 77.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 111.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 99.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 108.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 92.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 118.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 103.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 108.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 95.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 100.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 88.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 93.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 102.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 101.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 89.67 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 113.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 121.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 121.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 91.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 98.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 95.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 88.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 119.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 118.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 115.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 108.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 87.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 76.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 86.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 88.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 78.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 81.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 71.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 106.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 94.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 88.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 111.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 102.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 90.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 94.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 83.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 90.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 98.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 98.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 87.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 111.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 119.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 118.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 89.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 95.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 92.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 86.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 117.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 109.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 116.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 112.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 106.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 84.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 73.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 94.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 84.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 85.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 76.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 79.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 69.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 103.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 92.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 99.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 85.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 109.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 96.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 100.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 88.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 91.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 81.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 92.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 100.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 100.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 89.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 121.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 120.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 109.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 91.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 94.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 87.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 119.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 118.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 114.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 107.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 85.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 75.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 96.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 85.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 87.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 77.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 78.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 70.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 104.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 94.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 100.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 87.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 110.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 98.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 101.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 90.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 93.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 83.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 97.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 94.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 118.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 114.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 96.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 93.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 116.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 113.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 90.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 82.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 83.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 79.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 104.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 103.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 98.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 100.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 92.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 89.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 112.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 109.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 90.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 87.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 111.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 108.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 83.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 77.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 77.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 74.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 98.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 98.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 92.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 94.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 89.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 86.53 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 110.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 106.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 88.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 85.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 108.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 105.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 81.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 74.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 75.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 71.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 95.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 95.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 89.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 92.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 92.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 88.53 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 112.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 108.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 90.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 87.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 110.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 107.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 82.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 76.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 77.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 73.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 97.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 96.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 91.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 93.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 103.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 100.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 112.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 112.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 99.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 97.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 123.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 121.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 132.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 132.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 119.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 117.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 101.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 99.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 108.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 106.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 97.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 96.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 132.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 121.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 119.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 129.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 126.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 118.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 116.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 97.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 93.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 84.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 81.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 108.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 96.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 100.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 89.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 92.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 88.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 80.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 78.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 119.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 106.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 111.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 107.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 96.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 92.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 122.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 108.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 113.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 100.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 106.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 103.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 89.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 95.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 106.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 106.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 93.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 91.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 117.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 115.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 126.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 126.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 111.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 95.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 93.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 100.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 92.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 90.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 127.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 116.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 113.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 123.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 120.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 112.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 110.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 90.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 86.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 79.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 75.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 101.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 91.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 93.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 83.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 85.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 82.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 74.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 72.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 113.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 101.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 105.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 101.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 91.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 87.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 116.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 103.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 107.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 95.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 99.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 96.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 86.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 84.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 95.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 92.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 103.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 104.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 91.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 89.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 115.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 113.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 124.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 124.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 111.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 109.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 93.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 90.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 100.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 98.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 89.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 88.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 124.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 113.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 111.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 120.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 118.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 109.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 108.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 88.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 84.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 76.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 73.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 99.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 88.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 90.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 81.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 82.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 79.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 72.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 70.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 110.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 99.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 102.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 98.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 89.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 85.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 113.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 101.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 104.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 93.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 97.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 94.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 84.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 82.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 97.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 95.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 105.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 106.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 93.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 91.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 117.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 115.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 126.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 126.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 113.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 112.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 95.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 93.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 102.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 100.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 92.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 90.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 127.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 115.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 113.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 123.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 120.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 112.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 110.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 90.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 86.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 79.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 75.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 101.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 90.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 93.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 83.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 82.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 80.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 74.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 72.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 112.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 101.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 104.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 101.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 91.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 87.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 115.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 103.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 107.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 95.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 96.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 94.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 86.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 84.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 184.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 179.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 181.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 178.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 150.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 167.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 145.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 163.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 171.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 167.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 168.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 165.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 136.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 155.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 131.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 151.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 165.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 161.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 163.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 159.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 132.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 149.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 126.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 146.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 187.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 183.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 185.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 181.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 154.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 143.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 148.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 139.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 174.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 170.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 172.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 168.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 139.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 129.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 135.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 125.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 169.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 164.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 166.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 162.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 135.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 125.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 130.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 121.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 81.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 68.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 75.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 75.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 66.42 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 86.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 74.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 80.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 80.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 71.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 77.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 67.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 74.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 70.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 65.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 82.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 73.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 79.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 76.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 71.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 72.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 63.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 64.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 56.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 74.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 63.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 64.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 55.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 61.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 53.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 76.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 67.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 68.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 60.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 78.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 67.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 68.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 59.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 65.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 57.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 79.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 66.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 73.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 72.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 64.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 84.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 71.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 78.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 77.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 69.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 74.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 65.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 71.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 68.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 63.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 79.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 71.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 77.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 73.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 68.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 70.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 61.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 61.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 53.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 72.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 60.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 62.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 53.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 58.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 50.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 73.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 65.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 66.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 58.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 76.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 65.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 65.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 57.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 62.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 55.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 78.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 65.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 72.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 71.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 62.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 83.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 70.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 77.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 76.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 68.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 73.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 64.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 70.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 67.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 62.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 78.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 69.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 76.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 72.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 67.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 69.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 59.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 60.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 52.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 71.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 59.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 60.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 52.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 57.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 49.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 72.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 64.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 64.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 57.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 75.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 64.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 64.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 56.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 61.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 53.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 81.42 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 66.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 75.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 74.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 64.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 86.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 73.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 80.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 79.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 71.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 77.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 66.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 73.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 68.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 63.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 81.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 72.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 78.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 75.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 70.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 70.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 62.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 62.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 54.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 72.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 61.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 62.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 54.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 59.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 51.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 74.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 66.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 67.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 59.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 76.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 66.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 66.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 58.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 63.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 56.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 84.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 67.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 74.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 65.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 89.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 73.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 79.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 71.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 79.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 66.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 73.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 65.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 84.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 72.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 78.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 70.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 70.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 63.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 62.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 55.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 69.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 60.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 60.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 53.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 74.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 68.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 66.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 60.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 73.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 65.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 64.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 58.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 82.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 65.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 72.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 63.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 87.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 71.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 77.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 68.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 77.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 64.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 70.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 62.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 82.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 70.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 76.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 68.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 68.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 60.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 60.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 52.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 67.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 58.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 58.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 50.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 72.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 66.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 64.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 58.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 71.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 63.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 62.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 56.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 81.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 64.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 71.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 62.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 86.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 69.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 76.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 67.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 76.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 63.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 69.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 61.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 81.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 69.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 75.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 66.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 67.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 59.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 58.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 51.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 66.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 56.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 56.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 49.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 71.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 64.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 63.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 57.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 70.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 62.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 60.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 54.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 83.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 67.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 73.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 65.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 88.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 72.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 78.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 70.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 79.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 64.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 72.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 62.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 84.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 71.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 77.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 69.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 68.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 61.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 60.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 53.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 68.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 58.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 58.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 51.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 72.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 66.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 64.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 59.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 72.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 64.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 62.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 56.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 67.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 65.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 72.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 70.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 66.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 64.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 71.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 69.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 61.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 54.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 59.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 53.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 65.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 60.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 63.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 58.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 64.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 63.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 70.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 68.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 64.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 62.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 69.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 67.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 59.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 52.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 57.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 50.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 63.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 58.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 61.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 56.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 63.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 61.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 69.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 67.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 61.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 68.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 66.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 57.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 51.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 55.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 49.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 62.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 56.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 59.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 54.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 67.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 65.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 72.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 70.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 64.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 62.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 71.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 69.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 60.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 53.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 58.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 51.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 64.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 58.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 62.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 56.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 88.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 72.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 70.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 79.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 80.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 69.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 68.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 94.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 77.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 75.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 84.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 85.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 75.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 73.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 83.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 71.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 69.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 78.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 76.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 69.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 67.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 89.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 76.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 74.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 83.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 81.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 74.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 72.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 79.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 70.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 70.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 66.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 59.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 55.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 78.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 67.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 71.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 60.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 66.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 63.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 56.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 53.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 83.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 74.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 74.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 70.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 63.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 60.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 82.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 71.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 75.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 64.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 70.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 67.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 60.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 58.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 86.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 69.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 67.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 77.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 77.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 67.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 66.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 91.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 75.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 73.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 82.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 82.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 72.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 71.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 81.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 69.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 67.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 76.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 73.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 66.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 65.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 87.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 74.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 72.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 81.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 79.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 72.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 70.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 76.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 68.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 67.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 64.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 57.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 53.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 76.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 65.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 69.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 57.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 63.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 61.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 53.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 51.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 80.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 72.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 71.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 68.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 61.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 58.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 80.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 69.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 72.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 62.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 67.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 64.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 58.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 55.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 85.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 68.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 66.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 76.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 76.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 66.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 64.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 90.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 74.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 72.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 81.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 81.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 71.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 70.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 80.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 68.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 65.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 74.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 72.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 65.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 64.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 85.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 73.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 71.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 80.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 77.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 70.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 69.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 75.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 66.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 66.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 62.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 55.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 52.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 75.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 63.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 67.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 56.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 62.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 59.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 52.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 50.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 79.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 71.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 70.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 66.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 60.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 57.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 79.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 68.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 71.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 60.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 66.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 63.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 57.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 54.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 88.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 70.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 70.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 80.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 80.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 68.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 68.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 93.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 77.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 75.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 84.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 84.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 74.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 73.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 84.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 69.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 67.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 78.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 73.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 67.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 65.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 89.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:30 76.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 74.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:40 83.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 81.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:35 74.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 72.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 78.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 69.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:31 66.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 63.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:37 58.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 55.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:36 77.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:33 66.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 67.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 59.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:39 63.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 61.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 55.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 53.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 82.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 74.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 71.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 67.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 63.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 59.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:29 81.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:32 70.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 71.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 63.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 16-Jul-2026 09:34 67.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 65.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 16-Jul-2026 09:38 59.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 57.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 102.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 77.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 85.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 73.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 96.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 87.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 100.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 74.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 83.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 71.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 94.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 85.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 99.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 73.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 82.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 70.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 93.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 84.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 103.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 76.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 85.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 73.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 95.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 87.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 93.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 86.07 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 91.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 83.82 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 89.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 81.70 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 90.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 82.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 93.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 87.06 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 91.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 84.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 89.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 82.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 90.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 82.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 100.47 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 93.00 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 98.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 90.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 96.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 88.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 98.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 90.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 105.39 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 92.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 97.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 84.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 93.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 80.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 95.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 81.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 105.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 99.89 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 98.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 92.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 94.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 88.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 95.39 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 89.39 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 112.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 98.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 104.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 91.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 100.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 87.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 103.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 89.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 88.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 82.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 84.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 76.61 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 82.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 74.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 83.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 75.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 87.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 83.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 84.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 77.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 82.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 75.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 83.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 76.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 95.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 88.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 91.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 83.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 89.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 81.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 91.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 83.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 96.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 88.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 94.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 86.53 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 92.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 84.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 93.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 85.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 95.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 73.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 80.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 69.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 97.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 88.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 72.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 78.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 68.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 90.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 77.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 69.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 67.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 58.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 74.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 65.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 64.65 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 56.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 79.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 72.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 93.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 71.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 78.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 67.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 95.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 85.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 70.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 76.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 66.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 88.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 74.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 67.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 64.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 56.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 72.45 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 63.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 62.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 54.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 76.95 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 70.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 91.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 70.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 77.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 65.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 93.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 84.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 68.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 74.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 64.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 86.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 72.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 65.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 62.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 55.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 69.45 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 62.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 60.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 53.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 74.20 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 68.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 93.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 71.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 78.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 67.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 95.70 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 86.24 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 70.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 76.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 66.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 88.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 73.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 67.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 63.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 56.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 71.07 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 63.65 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 61.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 54.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 75.95 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 69.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 73.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 68.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 72.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 68.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:40 65.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 58.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 96.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 89.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 63.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 56.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 71.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 66.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 69.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 66.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 63.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 56.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 94.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 86.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 61.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 54.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 69.71 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 65.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 68.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 64.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 61.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 55.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 92.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 84.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 59.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 52.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 71.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 67.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 70.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 66.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 63.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 56.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 93.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 85.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 61.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 54.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 76.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 71.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 74.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 70.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 71.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 60.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 103.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 95.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 67.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 58.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 74.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 69.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 72.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 68.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 68.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 58.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 101.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 93.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 65.50 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 56.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 72.71 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 67.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 71.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 67.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 66.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 56.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 99.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 91.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 63.00 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 54.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 75.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 70.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 73.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 69.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 66.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 58.86 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 101.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 93.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 64.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 56.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 106.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 77.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 85.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 71.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 100.89 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 92.47 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 104.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 74.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 82.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 69.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 99.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 89.97 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 102.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 73.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 81.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 68.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 96.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 88.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 105.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 75.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 83.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 71.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 98.89 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 90.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 108.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 94.50 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 100.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 87.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 96.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 82.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 98.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 84.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 98.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 94.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 96.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 93.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 108.53 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 102.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 90.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 82.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 87.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 79.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 91.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 87.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 89.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 86.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 101.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 95.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 81.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 75.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:30 79.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 73.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 86.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 83.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 85.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 82.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 97.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 90.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 77.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 71.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 75.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 68.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 89.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 85.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 87.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 83.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 98.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 91.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 79.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 72.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 76.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 69.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 101.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 97.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 99.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 96.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 115.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 101.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 94.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 83.86 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 90.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 81.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 94.71 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 90.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 92.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 89.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 107.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 94.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 87.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 76.36 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 84.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 73.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 90.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 86.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 88.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 84.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 103.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 89.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 83.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 71.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 80.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 69.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 92.71 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 88.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 90.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 87.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 106.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 91.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 82.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 73.61 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 79.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 71.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 87.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 82.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 84.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 76.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 82.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 74.36 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 83.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 75.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 90.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 72.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 78.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 70.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 92.84 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 85.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 71.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 77.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 69.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 87.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 74.97 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 67.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 66.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 59.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 71.07 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 64.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:29 64.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 57.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 77.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 69.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 86.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 67.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 74.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 65.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 88.34 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 80.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 66.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 73.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 64.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 82.92 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 70.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 63.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 62.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 55.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 67.57 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 60.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 60.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 53.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 72.84 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 65.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 84.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 66.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 72.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 64.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 86.59 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 79.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 65.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 71.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 63.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 81.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 68.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 61.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:34 60.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 53.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 65.20 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 58.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 58.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 51.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 70.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 63.67 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 86.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 68.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 74.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 66.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 88.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 81.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 67.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 73.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 65.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 83.17 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 69.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:36 63.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 62.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 55.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:32 66.95 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 60.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 59.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:33 53.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 72.09 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 65.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 71.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 69.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 70.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 69.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 64.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 59.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 87.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 82.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 62.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 57.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 67.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 65.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 66.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 64.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:31 61.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 55.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 84.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 76.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 59.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 53.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 65.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 63.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 64.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 62.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 59.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 53.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 82.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 74.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 57.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 51.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 67.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 65.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 67.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 65.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 61.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 55.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 83.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 75.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:33 59.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 53.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 74.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 72.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:40 73.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 71.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 70.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 59.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 94.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 87.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:35 67.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 57.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:35 70.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 68.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 69.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:29 67.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:37 66.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 56.24 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 91.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 83.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 63.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 54.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:37 68.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 66.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 67.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 65.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 63.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 54.24 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 88.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 81.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:39 60.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 52.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:31 71.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 69.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 70.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 68.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:36 64.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:34 56.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 91.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 83.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 16-Jul-2026 09:38 62.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 54.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 104.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 75.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:32 84.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 72.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 98.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 89.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:38 99.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 71.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:38 79.65 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 68.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 95.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:30 85.61 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 98.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 69.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 78.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:30 66.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 92.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:34 83.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 100.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:39 72.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 81.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 16-Jul-2026 09:36 69.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 16-Jul-2026 09:32 95.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 85.99 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 105.54 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 94.05 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 109.53 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 98.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 105.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 94.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 109.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 98.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 106.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 95.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 109.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 99.11 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 116.03 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 104.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 119.61 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 108.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 127.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 116.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 131.14 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 120.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 122.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 110.99 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 125.99 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 114.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 118.25 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 106.98 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 121.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 110.81 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 119.72 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 109.03 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 123.28 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 112.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 127.96 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 117.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 131.72 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 121.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 123.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 111.77 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 126.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 115.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 118.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 108.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 122.41 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 111.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 120.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 109.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 124.01 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 113.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 138.90 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 128.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 142.44 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 131.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 133.76 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 122.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 137.43 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 126.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 129.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 118.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 133.25 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 122.20 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 131.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 120.99 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 135.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 124.71 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 129.62 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:37 124.32 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:29 121.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:39 122.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 105.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 94.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 109.15 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 98.18 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 106.15 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 95.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 110.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 99.11 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 101.27 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 86.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:35 103.98 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 89.96 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 100.93 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 89.82 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 104.03 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 92.77 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 111.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 97.40 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 114.60 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 100.67 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 141.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 122.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 144.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 125.09 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 122.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 106.40 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 125.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 109.19 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 114.65 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 100.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 117.65 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:31 103.24 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 115.07 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 101.20 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 118.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 104.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 142.21 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 127.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 145.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 130.00 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 123.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 111.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 126.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 114.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 115.50 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 104.07 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 118.47 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 107.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 115.33 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 103.92 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 118.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 107.00 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 153.03 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 133.59 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 156.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 136.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 134.39 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 117.69 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 137.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 120.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 126.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 111.74 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 129.41 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 114.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 126.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 113.19 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 129.84 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 115.98 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 100.98 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 86.96 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 104.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 89.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 101.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 89.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 104.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 92.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 110.00 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 98.27 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 112.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 101.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 109.69 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 101.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 112.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 104.37 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 160.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 151.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 163.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 153.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 137.72 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 124.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 140.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 127.64 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 123.82 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 113.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 127.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 116.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 159.47 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:39 153.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 162.44 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 156.43 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 139.37 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 128.68 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 142.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 131.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 125.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 116.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 127.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 119.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 100.50 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 88.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 106.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 94.14 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 100.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 89.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 106.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 95.06 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 111.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 99.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 116.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 105.13 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 116.40 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 104.53 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 122.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 110.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 114.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 102.26 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 119.35 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 107.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 112.50 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 100.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 117.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 106.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 113.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 101.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 119.23 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 107.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 116.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 105.38 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 122.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 111.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 114.56 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 103.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 119.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 108.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 113.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 101.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 118.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 107.27 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 114.57 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 102.64 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 119.79 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 108.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 127.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 115.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 134.01 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 121.79 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 125.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 113.36 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 131.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 119.01 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 123.92 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 111.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 129.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 117.61 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 126.71 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 114.09 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 132.06 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 119.74 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:31 120.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:33 118.20 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:35 116.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:40 118.82 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 104.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 88.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 109.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 94.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 104.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 89.37 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 110.28 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 95.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 114.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 99.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 120.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 104.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 125.92 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 110.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 131.54 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 115.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 120.13 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 104.93 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 125.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 110.38 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 116.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 101.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 122.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 106.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 126.11 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 110.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 132.09 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 116.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 119.60 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 105.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 125.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 110.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 116.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 102.15 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 122.47 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 107.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 137.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 121.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 142.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 127.08 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 131.67 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 116.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 137.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 121.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 128.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 112.41 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 133.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 118.20 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:30 107.56 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 96.38 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 112.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 101.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 108.36 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 97.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 113.71 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 102.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 112.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:32 99.61 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 116.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 103.23 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 113.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 101.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 117.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 105.70 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 101.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 90.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 105.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 94.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 102.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:29 112.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:37 101.16 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:36 141.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:40 129.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:37 114.28 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:34 103.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 91.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:29 106.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:32 115.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:31 104.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:30 144.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:39 133.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:36 118.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:34 106.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.32 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 116.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 105.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 118.33 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.66 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 122.01 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 115.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 103.85 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 119.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 107.71 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.83 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 102.32 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:34 117.51 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:33 106.17 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 116.20 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 105.01 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.88 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 108.87 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:32 118.71 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 107.88 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 122.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 111.91 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:36 115.88 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:29 105.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 119.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 108.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 114.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:31 103.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 107.57 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 116.70 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 120.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:36 109.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:40 130.28 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:35 119.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 133.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:39 122.04 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:37 127.33 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 116.15 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.48 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:38 119.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 125.79 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 114.74 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:37 117.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 117.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:39 132.02 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 120.64 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 102.08 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 90.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:31 105.76 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 94.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:35 102.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 112.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 141.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:36 129.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:31 114.41 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:33 103.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:30 91.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 106.51 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:37 116.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 104.92 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 145.07 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:40 133.90 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:40 118.18 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:32 106.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 95.57 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 95.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 84.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 98.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 87.11 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 97.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 85.42 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 99.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 88.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 108.00 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 95.63 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.67 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 98.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.61 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 106.67 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 120.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 108.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 111.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 100.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 113.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 102.36 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 97.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 110.89 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:35 99.29 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 111.41 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.02 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 114.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 121.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.43 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 123.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 110.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 112.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.01 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 115.79 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 103.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 109.88 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 98.07 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.58 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 100.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 111.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.70 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:36 114.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 132.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:30 118.47 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 135.31 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 120.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.60 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:33 111.87 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 127.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 114.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 121.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 108.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.02 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 111.44 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 123.73 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 111.11 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:40 126.37 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 113.74 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 97.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 84.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 98.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 87.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:30 97.48 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:34 85.55 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 100.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:38 88.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 105.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 93.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:32 108.75 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:40 95.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:38 107.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 94.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:37 97.06 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 135.87 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 123.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:31 138.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:29 125.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:39 123.79 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 111.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 126.18 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 113.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.91 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:29 121.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 109.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 137.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 123.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:33 140.17 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 126.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:34 125.31 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.14 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:32 115.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.44 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:36 107.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:37 123.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 110.66 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 95.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 83.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 99.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:34 87.57 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:38 96.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:32 111.68 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 89.00 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:36 139.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 117.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:31 113.68 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 90.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 84.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 100.46 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 112.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:31 92.58 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:40 141.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:30 121.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 114.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:31 94.64 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:40 88.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:33 106.20 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 94.46 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.48 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 99.64 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.46 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.61 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.40 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 107.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.44 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 99.41 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.13 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.25 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.72 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.63 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.87 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 104.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.63 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 113.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 101.82 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 108.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.29 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.74 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 97.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 101.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.76 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.47 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.92 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 119.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 108.06 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.26 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.90 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 110.11 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.56 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 111.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 89.00 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 139.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 117.91 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 113.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:23 90.94 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 112.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 92.70 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 141.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:23 121.51 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 114.58 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 94.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 99.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 87.36 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 103.28 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 91.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 99.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 88.04 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 103.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 92.02 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 107.66 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 94.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 96.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.96 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.67 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 103.25 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 91.83 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.56 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 103.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 114.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 102.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 143.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:23 131.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 116.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:23 104.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 92.73 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 107.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 118.17 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 106.54 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 147.14 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 135.64 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 120.25 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 108.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 96.77 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 102.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 117.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 119.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 104.98 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.63 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 108.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.57 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.09 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 117.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.26 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.21 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.12 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.29 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.00 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 113.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 117.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.18 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.22 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 110.23 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 104.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.68 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 108.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.16 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 122.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.97 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 131.87 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 120.27 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 134.90 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 123.47 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 117.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 132.07 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 120.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 116.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.53 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 119.24 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.44 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 118.88 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 133.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 122.07 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 104.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 114.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 102.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 143.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 131.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 116.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 104.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 92.85 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 107.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 118.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 106.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 147.14 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 135.64 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 120.25 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 108.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 96.77 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 96.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 85.54 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 99.77 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 87.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 98.29 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 86.24 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 100.99 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 88.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 109.06 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 96.34 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.73 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.17 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 119.55 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.49 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 122.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 109.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.54 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 97.97 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 111.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.98 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 112.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.28 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.85 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 122.35 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 108.12 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 110.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.02 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 101.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.73 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 104.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 98.89 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.43 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 100.28 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.16 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 133.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 119.16 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.44 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 121.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 115.08 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 109.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.21 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 112.13 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 111.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 127.43 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 114.31 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.72 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 94.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.56 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 96.55 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.34 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 95.07 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 97.63 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 123.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 139.31 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 125.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.74 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.24 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 114.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.87 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 107.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 109.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 138.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 124.36 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 141.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 126.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 113.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.09 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 115.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 108.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 124.21 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 111.23 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 100.16 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 84.03 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 103.75 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 88.02 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 100.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 101.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 90.43 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 130.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 119.21 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 103.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 92.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 84.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 104.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 105.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 94.27 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 134.36 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 122.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 107.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 96.33 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 88.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 94.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.32 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.61 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.71 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 99.83 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 120.15 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.15 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 117.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.99 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 99.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 113.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 97.44 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 101.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 115.98 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 104.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.88 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 98.14 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 102.13 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 100.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 112.74 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.17 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.34 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 102.16 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.12 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 131.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.63 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.02 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 128.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 108.10 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.43 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.15 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.08 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 113.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 101.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 90.55 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 130.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 119.34 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 103.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 92.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 105.82 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 94.39 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 134.36 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 122.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 107.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:23 96.33 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 107.82 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 96.21 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.06 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.49 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 125.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 114.23 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 154.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:23 142.92 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 127.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 116.17 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.23 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.38 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 129.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 117.88 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 157.90 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 147.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 131.00 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:23 119.94 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 118.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.83 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 110.81 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.30 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 116.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 121.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.95 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 123.66 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.22 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.64 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.77 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 114.50 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 124.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.51 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 117.42 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.69 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.79 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 114.61 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.24 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.08 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.61 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.32 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.43 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 124.66 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 139.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 127.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 133.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 121.96 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.51 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 125.03 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 131.94 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 120.30 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 134.97 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 123.50 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 135.01 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 123.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 138.04 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 126.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 125.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:23 114.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 154.28 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 142.92 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 127.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 116.17 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 112.50 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 129.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 118.00 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 158.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 147.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 131.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 119.94 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 101.33 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 89.90 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 104.18 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 92.04 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 102.70 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 90.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 105.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 93.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 100.57 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.14 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 103.41 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.95 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 111.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 113.74 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.95 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 105.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.29 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 104.23 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.88 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 104.07 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.69 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 106.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 115.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 118.42 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 106.07 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 108.61 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 103.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 117.93 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 105.66 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 104.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.57 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.30 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 137.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 123.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 140.77 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 126.03 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.07 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 116.68 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 132.70 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 119.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 113.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 116.37 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.21 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 116.04 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 131.84 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 118.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 98.60 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.91 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 100.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 99.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 115.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 141.93 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 128.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 144.30 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 130.11 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.85 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 116.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 132.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 118.82 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 124.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 127.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 114.10 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 142.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 128.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 145.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 131.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.72 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 117.50 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 133.43 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 120.10 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.84 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 115.38 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 95.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 83.88 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 99.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 87.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 96.18 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 105.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 93.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 134.19 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 122.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 107.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 95.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 84.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 100.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 109.37 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:23 97.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 138.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 126.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:23 111.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:23 99.51 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 88.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 94.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.57 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 99.79 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.64 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 113.25 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.96 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 107.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.71 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 99.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.26 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.28 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.23 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.25 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.72 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.87 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 104.77 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.76 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 97.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 102.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.22 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.24 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.74 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 98.01 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.64 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 102.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.14 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 127.76 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.18 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.67 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.64 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.81 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 111.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 121.90 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.38 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 113.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 105.61 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 93.74 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 134.31 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 122.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 107.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 95.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 109.49 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 97.57 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 138.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 126.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 111.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 99.51 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 102.20 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 90.88 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 105.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 94.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 102.88 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 112.45 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 141.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 129.97 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 114.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 103.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 91.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 116.10 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 104.92 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 145.08 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:23 133.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 118.18 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 106.98 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.57 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 105.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.70 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.38 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.75 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 104.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.43 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 108.09 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.21 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 102.70 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 117.89 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.58 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 105.39 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.13 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 109.24 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.09 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 108.25 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.84 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.26 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 105.44 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 118.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.95 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.96 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 120.84 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 110.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 119.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 133.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 122.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.71 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 116.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.87 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 119.73 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.00 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.33 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 118.20 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.24 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 117.83 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 132.40 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 121.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 103.00 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 112.58 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 101.29 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 141.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 129.97 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 114.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 103.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 91.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.89 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:23 116.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 105.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 145.08 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 133.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 118.18 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:23 106.98 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 95.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 96.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 85.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 99.09 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 87.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 97.60 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 85.80 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 100.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 88.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.38 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 95.89 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.92 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 98.73 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.86 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 107.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.36 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 109.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.85 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 100.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 114.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 97.65 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.27 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 99.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.79 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.39 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 101.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.80 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.19 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 110.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 101.39 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.04 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 104.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.25 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 98.45 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.84 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 101.11 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 111.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.96 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 132.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 118.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 135.69 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 121.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.85 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 114.76 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 121.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 109.18 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.40 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 111.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 111.49 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 126.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 114.00 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.03 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 94.11 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.01 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 96.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 107.65 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 94.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 97.32 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 123.60 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 138.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 125.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.04 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.31 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 126.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 114.33 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.68 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 109.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 137.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 124.01 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 140.43 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 126.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.69 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 113.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.40 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 115.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.81 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 108.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 123.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 110.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 96.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 84.47 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 100.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 88.45 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 97.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 112.56 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:23 89.00 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 140.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 117.78 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 114.44 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 90.94 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 85.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 101.21 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 113.10 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:23 92.71 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 142.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 121.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 115.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 94.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 89.13 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 107.08 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 99.33 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.24 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.40 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 117.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 104.37 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.13 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.57 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 108.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.20 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.01 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.86 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.60 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 101.39 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 117.75 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 105.50 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 110.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.71 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 102.70 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.10 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 97.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.17 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 98.74 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 114.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 102.73 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 125.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 128.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 116.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 122.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.68 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 108.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.78 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 110.87 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.27 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 114.32 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 112.69 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 89.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:22 140.27 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 117.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 114.56 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:23 90.94 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 16-Jul-2026 09:22 113.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 16-Jul-2026 09:22 92.84 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 16-Jul-2026 09:23 142.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 16-Jul-2026 09:22 121.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 16-Jul-2026 09:22 115.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 16-Jul-2026 09:22 94.77 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.47 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 95.42 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.17 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 99.16 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 107.16 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.45 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.05 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.24 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 117.16 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 106.07 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.75 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.94 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.59 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 118.05 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 132.28 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 121.72 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.45 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 112.12 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.12 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.80 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.38 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 108.24 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.07 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.95 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 120.86 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 110.17 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 124.41 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 113.87 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 129.22 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 118.70 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 132.86 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 122.55 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.31 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.91 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.08 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 116.89 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.78 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.18 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.55 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.05 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.26 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.70 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.14 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 114.54 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 140.03 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 129.59 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 143.58 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 133.01 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 134.90 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 123.66 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 138.56 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 127.19 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.84 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 119.79 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 134.38 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 123.34 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 133.02 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 122.13 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.69 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 125.84 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 130.88 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 125.59 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 122.30 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 123.75 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 107.28 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.57 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.18 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.24 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 102.41 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 87.94 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 105.11 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 91.09 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 102.07 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 90.96 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 105.16 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 93.91 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 112.72 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 98.54 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 115.73 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.63 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 142.80 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 123.43 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 145.73 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 126.22 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.71 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 107.54 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.46 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 110.33 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.79 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 101.59 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.66 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 104.38 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.20 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.34 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 119.15 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 105.25 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 143.34 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 128.19 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 146.20 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 131.13 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.89 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.24 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.86 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 115.19 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.63 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 105.21 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 119.61 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 108.16 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.47 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 105.06 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.45 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 108.13 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 154.17 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 134.73 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 157.26 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 137.77 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 135.52 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 118.82 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 138.62 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 121.74 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.48 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 112.88 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.55 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 115.92 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.02 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 114.33 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 130.97 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 117.12 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 111.13 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 99.41 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.84 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.66 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 110.83 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 102.55 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.80 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 105.51 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 161.59 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 152.18 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 164.68 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 155.09 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 138.86 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 125.87 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:23 142.08 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 128.78 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 124.95 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 114.47 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.18 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 117.26 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 160.60 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 154.62 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 163.58 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 157.56 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 140.50 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 129.82 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 143.48 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:23 132.77 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.20 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 117.94 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.04 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 16-Jul-2026 09:22 120.88 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 102.14 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 90.34 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.27 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.34 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 102.55 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 91.19 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 109.61 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 98.27 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 112.80 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 101.45 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 119.71 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 108.34 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 118.04 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 106.17 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.66 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.69 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.69 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.90 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.55 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.91 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 114.14 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 102.49 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 121.02 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 109.38 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 115.57 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.52 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 122.44 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.53 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.16 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.02 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.08 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 114.53 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.20 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 104.74 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 123.09 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.75 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 114.66 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.22 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 121.55 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.35 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 116.21 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 104.41 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 122.87 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 111.54 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.58 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 117.27 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.50 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 124.28 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.10 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.00 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 133.38 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 121.49 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.56 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.59 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 131.84 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 119.97 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.35 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 115.73 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 134.42 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 122.23 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 122.26 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 119.96 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 118.56 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 16-Jul-2026 09:22 120.46 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.01 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 89.54 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 112.54 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 96.55 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 106.42 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 90.25 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 113.62 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 97.47 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 116.80 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 100.34 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 123.11 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 107.37 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 127.81 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 111.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 134.23 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 118.09 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 122.02 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 105.81 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.42 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 112.70 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 118.78 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 102.19 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 125.06 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 109.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 128.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 111.77 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 135.75 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 118.79 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 121.70 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 105.99 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 129.12 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.13 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 118.59 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 103.03 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 126.00 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 110.05 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 139.23 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 122.30 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 145.66 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 128.69 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 133.56 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:23 116.92 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:23 139.86 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 123.43 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 130.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 113.30 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 16-Jul-2026 09:22 136.62 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 16-Jul-2026 09:22 119.80 KB
LICENSE 16-Jul-2026 09:24 11.09 KB
Artifactory/7.117.25 Server