Index of sw-kernelinferencelibrary-public-generic-local/e3a8eba02eb19f4485652f84f5095524350246b5/fmha/trtllm-gen
Name Last modified Size
../
include/ 22-Jul-2026 16:22 -
checksums.txt 22-Jul-2026 16:30 3.37 MB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 91.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 78.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 86.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 84.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 99.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 88.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 96.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 93.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 84.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 85.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 77.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 83.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 79.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 73.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 94.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 87.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 93.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 83.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 77.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 69.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 73.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 81.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 72.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 72.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 63.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 66.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 58.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 75.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 79.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 69.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 88.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 78.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 78.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 70.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 73.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 65.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 88.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 76.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 84.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 72.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 97.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 86.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 93.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 91.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 82.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 82.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 75.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 81.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 77.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 91.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 85.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 87.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 81.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 75.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 67.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 71.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 60.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 79.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 69.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 70.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 61.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 64.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 56.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 81.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 73.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 77.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 67.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 85.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 76.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 76.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 68.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 71.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 63.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 75.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 83.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 81.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 70.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 96.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 85.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 92.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 90.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 80.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 81.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 73.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 79.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 76.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 70.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 90.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 83.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 90.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 85.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 80.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 74.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 66.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 69.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 78.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 68.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 69.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 60.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 62.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 55.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 80.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 76.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 66.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 84.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 75.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 75.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 67.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 69.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 62.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 89.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 84.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 82.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 72.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 98.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 94.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 92.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 82.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 83.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 75.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 81.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 78.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 72.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 85.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 92.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 87.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 81.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 75.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 67.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 70.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 60.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 78.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 69.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 69.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 61.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 64.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 56.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 80.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 73.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 67.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 85.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 76.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 76.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 68.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 70.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 63.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 94.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 78.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 85.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 73.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 87.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 94.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 83.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 76.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 82.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 92.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 82.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 76.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 68.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 63.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 76.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 65.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 60.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 82.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 81.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 75.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 83.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 81.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 72.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 70.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 75.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 83.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 71.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 100.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 85.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 92.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 81.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 86.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 74.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 80.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 94.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 84.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 90.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 80.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 73.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 69.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 66.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 61.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 74.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 69.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 63.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 58.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 79.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 78.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 73.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 71.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 80.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 78.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 69.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 90.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 74.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 81.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 70.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 91.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 80.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 84.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 79.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 69.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 93.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 83.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 89.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 79.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 68.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 65.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 60.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 73.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 62.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 57.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 78.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 77.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 71.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 79.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 77.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 68.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 67.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 92.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 76.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 83.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 100.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 93.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 81.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 86.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 75.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 80.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 71.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 95.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 84.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 90.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 80.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 73.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 69.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 66.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 61.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 74.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 69.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 62.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 58.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 79.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 78.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 72.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 80.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 78.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 69.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 77.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 73.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 87.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 83.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 76.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 86.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 82.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 67.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 63.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 64.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 60.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 74.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 73.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 71.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 70.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 75.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 85.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 80.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 74.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 84.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 80.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 65.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 61.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 62.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 58.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 72.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 69.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 74.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 69.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 84.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 79.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 69.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 82.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 78.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 64.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 60.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 61.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 57.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 71.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 70.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 68.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 67.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 76.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 86.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 81.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 74.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 80.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 65.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 61.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 62.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 58.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 72.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 69.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 97.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 82.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 80.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 90.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 89.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 77.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 76.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 92.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 90.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 98.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 87.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 85.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 80.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 79.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 84.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 76.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 75.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 100.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 90.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 89.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 97.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 94.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 86.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 85.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 84.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 76.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 76.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 72.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 65.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 62.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 86.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 76.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 77.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 68.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 71.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 68.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 61.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 59.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 90.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 82.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 83.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 79.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 72.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 83.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 83.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 75.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 75.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 68.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 66.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 95.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 80.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 78.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 88.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 87.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 75.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 73.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 90.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 88.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 97.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 96.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 85.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 83.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 78.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 77.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 85.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 82.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 74.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 73.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 98.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 88.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 87.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 95.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 84.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 83.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 82.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 74.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 74.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 70.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 63.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 60.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 83.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 75.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 66.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 69.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 66.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 59.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 57.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 80.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 81.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 77.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 70.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 67.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 90.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 81.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 73.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 75.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 73.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 66.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 64.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 94.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 78.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 77.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 87.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 86.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 74.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 103.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 89.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 87.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 95.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 84.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 82.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 75.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 84.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 81.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 73.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 72.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 97.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 85.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 94.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 90.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 83.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 81.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 81.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 73.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 73.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 69.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 58.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 82.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 72.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 74.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 65.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 67.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 65.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 58.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 55.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 86.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 79.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 79.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 75.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 69.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 65.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 88.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 79.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 80.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 74.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 71.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 65.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 62.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 96.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 80.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 79.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 89.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 88.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 76.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 74.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 105.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 90.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 89.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 97.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 85.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 84.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 90.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 79.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 77.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 86.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 83.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 75.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 73.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 89.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 87.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 96.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 92.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 85.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 83.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 82.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 74.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 74.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 68.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 63.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 60.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 83.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 74.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 75.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 66.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 67.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 65.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 59.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 57.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 88.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 80.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 80.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 75.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 70.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 67.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 90.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 81.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 81.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 73.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 74.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 72.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 66.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 64.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 91.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 99.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 86.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 96.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 112.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 89.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 97.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 84.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 102.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 94.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 88.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 96.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 82.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 93.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 90.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 102.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 97.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 107.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 94.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 118.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 126.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 126.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 102.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 99.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 93.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 124.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 116.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 123.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 92.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 80.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 91.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 93.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 84.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 76.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 111.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 99.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 92.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 117.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 102.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 87.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 92.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 100.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 88.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 109.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 90.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 96.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 94.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 110.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 86.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 75.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 85.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 87.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 78.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 80.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 71.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 87.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 110.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 101.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 93.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 82.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 89.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 97.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 86.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 110.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 106.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 94.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 91.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 85.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 83.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 72.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 93.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 83.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 75.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 78.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 68.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 92.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 85.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 95.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 98.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 90.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 80.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 90.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 98.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 87.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 89.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 95.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 92.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 86.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 117.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 109.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 116.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 113.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 83.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 73.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 93.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 83.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 84.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 75.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 76.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 69.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 85.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 91.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 97.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 93.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 117.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 114.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 95.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 92.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 115.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 112.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 89.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 81.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 82.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 79.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 103.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 102.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 97.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 99.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 91.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 88.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 111.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 108.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 89.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 86.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 110.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 107.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 82.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 76.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 77.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 73.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 97.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 97.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 91.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 94.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 88.78 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 85.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 109.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 105.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 87.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 84.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 107.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 104.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 80.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 74.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 74.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 71.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 94.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 94.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 88.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 91.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 90.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 86.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 110.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 107.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 88.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 85.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 108.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 106.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 80.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 74.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 75.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 95.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 95.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 89.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 91.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 101.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 99.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 111.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 111.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 97.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 96.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 122.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 120.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 131.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 131.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 116.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 100.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 98.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 95.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 131.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 120.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 118.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 128.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 117.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 92.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 84.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 80.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 98.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 91.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 87.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 79.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 77.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 105.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 110.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 106.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 95.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 91.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 121.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 107.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 102.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 90.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 88.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 96.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 94.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 105.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 92.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 90.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 114.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 125.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 111.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 94.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 92.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 99.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 89.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 112.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 122.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 111.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 109.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 89.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 85.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 78.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 74.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 100.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 90.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 92.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 82.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 84.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 81.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 73.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 101.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 100.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 90.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 86.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 115.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 95.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 85.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 83.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 93.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 91.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 89.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 88.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 112.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 123.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 108.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 91.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 89.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 99.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 97.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 88.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 87.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 123.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 110.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 107.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 87.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 82.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 76.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 98.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 87.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 89.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 80.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 78.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 71.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 109.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 98.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 101.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 97.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 88.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 112.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 103.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 92.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 96.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 93.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 83.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 81.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.92 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 92.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 104.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 91.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 89.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 115.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 113.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 124.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 111.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 109.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 93.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 91.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 100.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 89.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 88.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 124.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 111.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 118.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 108.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 83.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 73.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 88.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 90.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 80.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 80.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 78.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 72.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 70.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 110.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 98.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 89.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 85.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 100.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 93.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 94.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 92.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 84.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 82.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 182.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 178.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 179.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 176.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 148.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 165.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 143.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 162.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 169.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 165.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 166.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 163.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 134.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 153.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 129.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 150.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 163.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 159.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 161.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 157.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 130.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 148.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 125.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 144.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 185.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 181.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 183.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 179.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 151.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 141.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 146.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 137.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 172.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 168.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 170.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 166.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 137.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 128.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 133.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 124.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 167.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 162.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 164.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 161.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 133.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 123.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 129.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 119.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 80.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 68.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 75.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 74.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 86.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 73.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 80.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 79.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 71.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 75.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 67.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 73.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 70.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 65.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 81.53 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 72.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 78.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 75.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 70.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 71.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 62.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 63.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 55.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 62.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 63.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 55.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 60.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 52.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 75.70 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 67.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 67.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 60.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 66.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 67.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 59.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 64.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 56.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 78.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 65.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 72.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 72.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 63.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 84.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 71.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 78.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 77.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 68.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 73.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 63.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 79.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 70.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 76.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 73.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 68.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 60.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 61.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 53.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 60.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 61.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 52.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 58.26 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 50.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 73.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 65.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 65.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 57.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 75.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 64.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 65.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 57.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 62.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 54.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 76.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 64.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 71.69 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 71.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 62.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 82.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 70.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 76.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 76.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 67.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 63.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 70.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 66.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 61.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 78.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 69.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 75.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 72.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 67.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 67.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 59.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 60.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 52.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 70.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 59.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 60.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 51.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 57.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 49.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 72.20 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 64.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 64.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 56.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 74.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 63.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 64.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 55.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 61.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 53.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 78.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 66.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 73.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 72.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 64.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 84.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 71.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 78.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 78.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 69.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 73.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 65.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 71.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 68.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 63.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 79.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 71.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 77.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 73.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 69.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 60.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 61.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 53.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 71.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 60.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 61.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 52.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 58.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 50.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 72.95 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 65.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 65.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 57.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 75.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 64.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 64.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 57.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 62.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 54.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 73.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 65.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 89.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 79.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 70.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 78.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 66.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 64.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 71.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 77.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 69.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 69.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 62.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 54.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 69.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 59.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 59.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 52.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 74.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 66.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 59.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 73.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 65.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 63.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 57.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 80.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 65.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 63.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 86.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 70.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 76.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 68.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 76.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 64.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 62.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 82.31 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 69.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 75.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 67.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 67.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 60.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 59.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 52.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 66.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 57.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 57.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 50.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 71.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 66.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 63.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 57.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 70.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 62.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 61.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 55.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 79.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 63.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 70.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 61.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 85.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 69.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 75.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 67.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 74.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 62.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 69.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 61.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 81.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 74.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 66.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 65.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 58.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 51.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 65.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 56.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 56.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 49.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 70.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 65.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 62.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 56.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 69.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 61.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 60.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 81.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 65.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 63.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 87.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 77.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 68.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 76.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 64.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 62.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 70.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 76.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 66.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 59.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 59.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 52.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 66.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 57.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 57.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 50.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 71.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 65.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 63.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 57.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 70.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 62.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 61.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 55.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 66.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 64.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 72.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 65.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 69.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 60.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 54.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 58.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 52.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 65.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 59.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 62.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 57.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 62.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.91 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 67.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 63.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 61.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 58.65 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 52.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 56.62 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 50.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 62.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 57.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 60.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 55.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 63.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 61.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 66.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 62.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 60.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 67.84 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 65.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 57.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 51.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 55.37 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 49.18 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 61.51 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 56.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 59.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 54.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 65.03 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 63.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 68.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 64.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 62.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 69.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 58.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 51.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 56.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 50.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 62.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 57.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 60.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 55.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 87.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 71.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 79.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 79.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 69.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 67.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 93.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 77.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 75.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 84.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 84.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 73.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 70.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 77.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 75.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 68.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 88.28 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 76.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 74.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 83.12 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 80.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 73.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 77.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 69.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 69.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 65.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 58.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 55.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 78.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 66.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 70.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 59.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 65.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 62.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 55.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 53.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 74.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 73.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 69.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 63.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 59.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 82.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 70.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 74.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 63.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 69.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 66.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 59.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 57.54 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 84.88 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 69.42 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 67.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 77.06 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 77.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 67.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 65.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 90.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 74.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 82.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 82.40 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 72.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 80.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 68.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 66.71 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 75.61 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 73.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 66.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 64.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 86.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 80.99 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 78.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 75.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 67.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 67.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 63.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 56.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 53.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 76.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 64.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 68.52 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 57.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 63.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 60.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 53.43 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 51.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 80.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 72.01 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 71.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 67.49 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 61.21 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 57.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 79.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 68.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 61.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 67.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 64.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 57.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 55.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 83.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 68.17 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 66.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 75.81 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 76.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 65.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 64.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 89.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 73.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.66 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 81.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 71.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 69.50 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 78.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 67.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 74.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 72.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 65.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 63.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 84.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 72.98 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 70.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 79.74 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 77.23 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 70.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 74.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 65.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 62.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 55.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 52.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 74.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 63.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 67.27 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 56.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 62.11 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 59.22 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 52.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 49.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 78.80 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 70.76 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 70.00 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 66.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 60.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 56.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 78.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 67.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 71.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 60.58 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 65.97 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 63.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 56.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 54.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 85.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 70.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.16 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 77.56 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 77.89 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 67.64 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 66.25 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 91.45 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 75.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 73.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 83.19 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 83.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 73.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 71.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 80.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 69.47 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 67.33 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 76.24 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 74.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 66.94 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 65.55 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 86.90 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 75.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 81.87 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 79.35 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 72.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 75.75 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 67.44 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 65.39 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 62.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 56.96 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 53.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 75.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 64.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 66.14 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 57.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 61.86 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 59.72 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 53.68 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 51.41 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 80.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 72.38 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 69.63 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 66.36 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 61.46 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 58.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 79.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 69.10 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 70.13 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 62.08 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.85 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 63.83 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 58.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 55.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.29 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 76.59 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 84.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 73.30 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.07 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 88.15 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 100.04 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 74.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 82.73 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.05 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 93.82 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 86.02 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 98.79 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 73.09 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 81.48 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 69.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 92.57 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.77 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.67 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 75.34 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 83.60 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.93 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 94.32 KB
fmhaSm100aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 86.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 93.01 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 86.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 90.76 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 83.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 88.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 81.86 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 89.01 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 82.06 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 92.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 86.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 90.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 84.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 87.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 81.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 88.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 82.03 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 101.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 95.28 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 99.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 92.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 97.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 90.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 98.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 91.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 105.01 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 93.39 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 95.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 85.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 91.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 80.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 92.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 80.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 104.78 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 100.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 96.53 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 91.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 92.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 87.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 92.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 86.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 113.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 100.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 105.42 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 91.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 101.75 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 87.01 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 102.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 87.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 87.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 81.61 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 83.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 76.03 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 81.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 73.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 82.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 74.40 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 86.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 81.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 82.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 76.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 80.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 74.29 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 81.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 74.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 95.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 90.04 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 92.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 85.29 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 90.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 83.17 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 91.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 84.16 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 95.76 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 88.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 93.51 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 86.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 91.26 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 84.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 91.76 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 84.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 95.31 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 73.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 80.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 97.17 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 88.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 77.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 89.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 77.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 69.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 66.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 58.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 73.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 65.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 64.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 56.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 79.09 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 71.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 93.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 78.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 66.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 95.17 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 86.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 75.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 65.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 87.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 74.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 67.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 64.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 56.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 71.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 63.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 61.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 76.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 69.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 91.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 69.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 76.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 93.42 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 84.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 73.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 64.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 85.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 71.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 66.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 61.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 54.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 68.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 61.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 59.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 52.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 73.67 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 67.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 70.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 77.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 66.39 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 94.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 86.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 69.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 74.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 65.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 87.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 72.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 66.61 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 62.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 55.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 69.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 62.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 60.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 53.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 74.42 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 68.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 71.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 67.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 65.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 58.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 95.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 88.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 62.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 56.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 66.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 65.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 63.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 56.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 92.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 86.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 60.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 54.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 69.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 64.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 64.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 60.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 54.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 90.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 84.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 58.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 52.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 70.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 66.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 69.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 65.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 62.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 55.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 91.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 84.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 59.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 53.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 75.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 74.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 70.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 60.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 104.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 97.78 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 67.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 57.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 73.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 69.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 68.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 58.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 102.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 95.53 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 64.97 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 55.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 67.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 66.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 65.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 56.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 100.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 93.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 62.47 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 53.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 73.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 68.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 65.00 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 57.21 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 101.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 94.28 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 62.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 107.33 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 76.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 101.58 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 92.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.46 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 74.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 82.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 69.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.33 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 90.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 103.71 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 80.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 67.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 96.43 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.21 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 74.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 82.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 97.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 89.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 107.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 95.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 98.63 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 87.47 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 94.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 82.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 95.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 82.83 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 98.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 94.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 93.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 107.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 103.00 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 89.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 82.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 87.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 80.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 90.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 88.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 85.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 99.28 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 93.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 80.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 75.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 78.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 82.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 84.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 81.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 95.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 89.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 76.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 70.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 74.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 83.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 84.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 81.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 95.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 89.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 77.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 74.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 101.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 98.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 99.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 96.70 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 116.67 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 102.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 93.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 84.46 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 89.97 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 82.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 93.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 89.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 91.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 88.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 108.29 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 93.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 86.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 75.58 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 83.22 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 73.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 89.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 87.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 84.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 104.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 89.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 82.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.08 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 79.22 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 90.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 86.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 88.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 84.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 105.25 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 89.72 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 80.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.58 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 77.72 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 69.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 87.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 81.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 83.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 76.03 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 81.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 73.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 81.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 74.40 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 89.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 78.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 76.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 86.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 73.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 66.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 65.65 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 59.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 70.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 63.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 63.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 56.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 75.06 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 85.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 87.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 80.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 66.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.49 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 64.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 82.39 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 70.18 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 62.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 62.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 54.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 67.04 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 59.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 59.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 52.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 71.81 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 64.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 83.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 65.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 72.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 63.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 86.31 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 78.48 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 64.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 70.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 62.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 80.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 67.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 60.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 59.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 53.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 64.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 58.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 57.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 51.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 69.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 63.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 85.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 73.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 65.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 79.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 66.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 72.12 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 64.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 82.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 67.90 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 61.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 60.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 53.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 65.41 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 58.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 58.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 51.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 70.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 63.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 69.27 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 64.40 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 59.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 86.44 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 80.94 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 62.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 57.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 66.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 64.77 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 65.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 63.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 61.15 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 82.69 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 75.91 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 58.99 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 52.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 64.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 63.14 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 62.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 59.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 53.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 80.57 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 73.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 56.87 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 51.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 66.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 64.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 65.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.07 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 60.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 53.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 81.19 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 74.29 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 57.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 51.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 74.05 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 72.02 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 73.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.45 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 69.50 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 59.46 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 95.56 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 89.67 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 66.60 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 57.30 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 69.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 67.52 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 66.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 65.88 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 55.71 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 92.06 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 84.92 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 62.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 53.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 67.93 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 65.89 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 67.10 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 65.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 63.38 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 53.71 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 89.81 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 82.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 60.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 51.68 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.55 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.64 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.73 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 66.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 63.13 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 54.96 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 91.06 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 83.79 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 60.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 52.80 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 75.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 83.74 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.32 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 98.28 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.58 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 98.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 79.24 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 67.82 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 94.66 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 85.83 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 97.35 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 69.23 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 77.62 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 66.20 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 91.75 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 83.83 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 98.85 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.98 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 79.37 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 67.95 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 93.00 KB
fmhaSm100aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 84.96 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:26 111.67 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:33 98.95 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:29 144.32 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:32 131.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:29 114.38 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:28 101.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:26 113.38 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:25 100.73 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:25 145.30 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:31 132.70 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:25 116.22 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:31 103.55 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:28 111.80 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:29 99.07 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:29 144.45 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:26 131.77 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:25 114.50 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:29 101.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:29 113.51 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:25 100.85 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:31 145.43 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:33 132.82 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:29 116.34 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:33 103.55 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:33 113.52 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:30 94.21 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:29 145.84 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:31 127.12 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:25 116.37 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:32 96.91 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:28 115.23 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:25 95.88 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:29 147.65 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:28 128.05 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:27 118.07 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:28 98.69 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:27 113.64 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:30 94.21 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:33 145.96 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:32 127.24 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:25 116.49 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:30 97.03 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:32 115.35 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:27 96.00 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:25 147.65 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:26 128.17 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:31 118.20 KB
fmhaSm100aKernel_QkInt8VE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:30 98.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 140.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 114.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 123.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 109.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.64 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 113.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 120.63 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 108.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 111.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 104.31 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 97.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 113.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 107.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 100.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 94.29 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 138.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 111.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 120.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 106.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 126.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 110.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 117.82 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 105.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.71 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 101.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 95.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 110.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 104.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 97.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 91.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 136.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 110.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 119.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 105.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 108.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 116.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 104.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 112.85 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 99.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 93.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 109.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 102.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 96.13 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 90.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 138.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 121.12 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 107.11 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 127.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 110.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 118.37 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 105.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.35 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 100.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 94.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 110.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 104.00 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 97.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 91.37 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 114.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 112.68 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 107.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 102.86 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 97.53 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 99.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 94.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 111.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 105.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 109.87 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 105.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 99.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 94.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 96.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 91.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 109.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 104.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.33 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 103.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 98.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 93.31 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 94.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 90.01 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 106.79 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 110.78 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 105.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 100.18 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 94.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 96.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 91.06 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 118.31 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 112.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 116.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 111.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 108.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 92.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 103.78 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 89.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 109.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 113.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 108.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 105.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 89.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 100.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 86.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 113.96 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 108.37 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 112.36 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 107.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 104.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 88.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 99.29 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 84.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 116.26 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 110.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 114.78 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 109.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 103.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 90.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 99.93 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 86.65 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 148.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 119.01 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 128.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 140.00 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 126.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 145.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 116.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 125.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 110.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 137.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 123.96 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 144.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 114.78 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 124.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 108.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 135.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 122.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 146.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 117.11 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 126.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 110.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 137.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 124.18 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 149.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 145.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 147.14 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 143.74 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 136.17 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 131.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 128.22 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 127.72 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.61 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 138.22 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.33 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 136.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 128.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 124.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 121.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 120.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 139.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 135.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 137.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 133.61 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 124.99 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 121.92 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 117.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 117.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 141.12 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 136.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 138.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 135.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 125.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 122.47 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 118.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 118.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 153.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 148.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 150.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 147.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 139.24 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 119.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 133.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 114.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 146.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 141.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 143.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 140.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 131.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 112.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 126.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 108.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 143.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 138.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 140.92 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 137.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 128.24 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 110.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 122.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 105.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 144.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 140.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 142.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 138.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 129.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 111.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 122.31 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 107.26 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 225.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 219.56 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 222.21 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 217.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 189.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 205.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 180.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 200.57 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 209.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 203.42 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 206.08 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 201.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 172.87 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 190.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 164.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 185.41 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 202.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 196.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 198.69 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 194.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 167.71 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 183.21 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 159.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 178.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 230.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 223.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 226.53 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 221.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 192.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 176.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 185.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 171.85 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 214.01 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 207.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 210.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 205.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 175.71 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 161.22 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 169.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 155.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 206.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 200.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 203.00 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 198.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 170.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 155.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 164.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 150.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 127.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 108.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 98.00 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 99.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 106.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 96.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.17 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 101.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 96.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 85.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 104.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 91.73 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 93.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 82.76 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.86 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 97.63 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 105.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 95.19 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 116.34 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 96.64 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 103.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 94.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 108.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 98.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 93.35 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 82.62 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 101.64 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 89.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 90.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 80.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 96.23 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 104.09 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 93.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 114.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 95.11 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 102.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 92.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.77 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 96.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 91.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 81.08 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 100.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 87.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 88.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 78.54 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.53 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.15 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 106.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 95.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 96.91 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 104.59 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 94.46 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.06 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 97.61 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 92.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 82.38 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 101.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 88.68 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 90.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 79.72 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 99.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 97.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 98.52 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 96.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 94.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 85.24 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 91.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 82.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.79 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 94.47 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 95.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 93.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 91.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 82.43 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 88.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 80.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 95.26 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 92.94 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 94.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 91.98 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 90.26 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 81.02 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 87.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 78.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 97.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 95.51 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.87 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 94.55 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 92.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 82.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 89.18 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 79.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 103.63 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 101.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.67 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.10 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 100.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 84.20 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 96.40 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 81.50 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 100.81 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 98.37 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 99.85 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 97.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 97.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 81.36 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 93.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 78.66 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 99.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 98.32 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 95.88 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 95.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 79.95 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 92.03 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 77.25 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 101.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 99.13 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 100.74 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 98.30 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 95.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 81.83 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 92.80 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 79.01 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 137.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 104.97 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 114.70 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 101.27 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 134.05 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.84 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 135.16 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 102.28 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 111.89 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 98.45 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 131.08 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.13 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 133.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 100.75 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 110.48 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 97.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 129.65 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 115.60 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 136.04 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 103.07 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 112.93 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 99.49 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 131.58 KB
fmhaSm100fKernel_QFp16KvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.48 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 105.19 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 94.00 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 109.62 KB
fmhaSm100fKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 98.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 105.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 93.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 110.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 99.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 104.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 93.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 109.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 99.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 115.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 104.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 120.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 110.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 138.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 123.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 133.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 125.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 116.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 148.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 127.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 138.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 128.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 120.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 129.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 130.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 115.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 136.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 125.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 135.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 118.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 128.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 124.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 127.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 114.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 136.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 121.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 115.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 115.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 135.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 127.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 132.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 120.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 140.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 105.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 132.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 115.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 125.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 143.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 131.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 123.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 124.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 121.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 110.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 131.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 127.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 106.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 116.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 98.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 121.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 109.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 113.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 109.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 95.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 106.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 129.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 120.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 126.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 115.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 132.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 113.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 99.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 110.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 96.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 117.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 140.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 114.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 120.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 107.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 121.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 117.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 107.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 102.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 113.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 110.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 106.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 103.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 94.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 118.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 106.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 122.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 105.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 101.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 126.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 113.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 98.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 123.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 111.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 127.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 95.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 104.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 91.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 130.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 111.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 121.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 104.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 141.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 126.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 109.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 108.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 129.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 123.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 111.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 104.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 110.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 118.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 106.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 121.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 92.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 126.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 123.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 111.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 127.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 113.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 95.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 104.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 141.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 121.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 130.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 114.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 148.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 132.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 120.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 128.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 114.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 137.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 120.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 105.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 130.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 114.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 113.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 100.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 131.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 119.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 136.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 114.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 122.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 107.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 144.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 127.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 113.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 120.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 106.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 132.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 121.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 106.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 112.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 97.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 121.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 106.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 104.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 92.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 126.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 114.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 132.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 109.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 117.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 140.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 123.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 107.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 116.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 101.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 103.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 107.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 92.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 115.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 101.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 98.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 87.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 122.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 133.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 109.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 118.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 103.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 141.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 124.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 108.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 117.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 101.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 129.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 103.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 106.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 92.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 115.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 101.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 98.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 87.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 111.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 126.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 115.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 131.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 120.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 120.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 109.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 125.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 114.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 117.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 106.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 122.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 111.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 117.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 106.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 112.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 139.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 95.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 107.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 90.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 128.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 93.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 98.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 89.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 128.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 116.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 93.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 80.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 100.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 87.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 96.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 85.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 90.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 77.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 145.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 138.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 132.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 120.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 153.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 132.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 143.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 134.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 124.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 136.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 126.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 134.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 121.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 141.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 130.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 139.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 123.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 136.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 120.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 128.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 110.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 138.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 126.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 140.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 120.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 106.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 104.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 140.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 126.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 132.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 143.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 131.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 144.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 130.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 109.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 109.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 140.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 130.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 127.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 148.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 135.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 129.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 131.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 126.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 117.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 136.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 123.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 132.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 119.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 130.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 120.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 132.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 121.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 131.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 118.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 115.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 101.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 109.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 96.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 134.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 121.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 106.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 137.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 126.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 137.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 123.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 136.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 115.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 123.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 144.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 130.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 126.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 111.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 113.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 122.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 113.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 106.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 133.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 117.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 127.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 110.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 126.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 114.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 98.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 129.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 117.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 126.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 112.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 97.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 104.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 91.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 131.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 120.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 102.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 134.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 122.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 131.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 117.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 114.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 100.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 95.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 138.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 115.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 126.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 124.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 146.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 131.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 127.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 129.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 123.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 114.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 134.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 128.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 117.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 127.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 130.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 118.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 126.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 112.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 98.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 104.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 132.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 119.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 135.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 123.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 131.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 118.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 101.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 109.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 95.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 149.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 126.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 136.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 119.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 148.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 132.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 131.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 144.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 118.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 128.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 111.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 142.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 141.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 113.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 123.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 106.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 139.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 122.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 142.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 114.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 124.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 107.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 142.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 124.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 105.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 93.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 110.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 99.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 104.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 93.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 109.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 99.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 101.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 86.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 104.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 89.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 100.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 89.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 104.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 92.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 111.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 97.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 114.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 100.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 146.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 151.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 163.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 137.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 144.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 152.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 152.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 165.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 139.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 145.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 133.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 148.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 158.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 126.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 142.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 138.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 149.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 160.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 128.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 143.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 141.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 132.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 122.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 115.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 145.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 130.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 145.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 127.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 109.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 135.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 122.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 144.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 135.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 124.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 122.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 149.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 134.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 163.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 149.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 138.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 131.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 136.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 147.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 122.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 137.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 139.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 151.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 124.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 132.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 133.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 142.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 127.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 124.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 136.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 147.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 113.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 130.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 116.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 106.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 100.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 133.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 117.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 146.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 132.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 110.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 93.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 125.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 109.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 137.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 151.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 135.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 96.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 124.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 124.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 128.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 140.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 122.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 131.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 132.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 145.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 118.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 111.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 126.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 135.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 104.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 120.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 118.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 130.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 141.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 124.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 114.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 109.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 100.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 126.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 139.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 102.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 100.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 117.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 103.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 101.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 130.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 144.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 130.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 91.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 118.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 104.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 124.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 128.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 140.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 131.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 132.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 145.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 125.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 111.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 125.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 135.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 130.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 140.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 124.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 114.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 100.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 137.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 121.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 101.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 99.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 116.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 102.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 101.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 129.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 143.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 127.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 90.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 117.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 104.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 142.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 126.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 145.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 130.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 123.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 111.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 126.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 114.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 115.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 103.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 118.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 107.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 114.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 102.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 117.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 105.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 164.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 118.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 127.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 135.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 114.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 152.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 115.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 124.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 124.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 113.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 152.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 141.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 116.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 103.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 125.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 111.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 124.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 110.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 111.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 100.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 154.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 154.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 168.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 143.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 147.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 156.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 156.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 169.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 144.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 148.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 141.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 152.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 163.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 132.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 146.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 143.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 154.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 165.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 133.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 147.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 153.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 140.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 133.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 123.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 149.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 134.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 166.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 149.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 132.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 140.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 155.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 139.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 136.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 153.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 138.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 169.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 153.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 130.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 117.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 143.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 127.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 139.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 140.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 152.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 127.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 134.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 141.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 144.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 156.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 129.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 136.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 126.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 138.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 147.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 132.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 128.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 141.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 151.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 118.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 135.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 134.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 124.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 117.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 108.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 137.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 118.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 152.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 135.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 99.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 128.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 137.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 120.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 111.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 141.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 156.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 139.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 101.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 131.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 113.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 132.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 133.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 145.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 120.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 126.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 135.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 137.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 150.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 130.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 131.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 140.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 109.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 124.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 122.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 135.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 145.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 128.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 126.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 111.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 112.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 143.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 128.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 93.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 120.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 129.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 120.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 114.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 106.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 134.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 149.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 133.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 96.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 124.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 132.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 145.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 125.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 135.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 137.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 150.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 123.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 129.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 130.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 140.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 122.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 134.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 146.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 112.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 128.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 125.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 111.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 128.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 143.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 93.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 117.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 103.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 128.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 114.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 148.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 132.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 96.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 121.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 100.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 86.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 104.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 89.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 101.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 89.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 104.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 92.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 160.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 167.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 155.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 154.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 158.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 164.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 153.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 131.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 120.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 139.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 127.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 136.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 124.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 127.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 118.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 109.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 97.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 111.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 100.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 108.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 100.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 111.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 103.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 153.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 154.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 154.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 152.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 159.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 150.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 150.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 145.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 153.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 146.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 162.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 153.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 153.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 144.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 151.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 140.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 134.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 136.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 131.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 123.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 118.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 130.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 117.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 139.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 135.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 126.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 126.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 132.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 123.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 112.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 126.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 115.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 113.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 120.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 184.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 193.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 179.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 170.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 182.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 182.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 190.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 178.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 158.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 147.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 168.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 155.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:30 172.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 165.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 149.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 154.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 154.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 145.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 158.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 149.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 161.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 152.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 138.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 128.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 141.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 131.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 124.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 115.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 127.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 118.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 161.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 160.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 159.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 158.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 159.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 152.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 158.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 151.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 156.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 148.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 156.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 144.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 136.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 131.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 138.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 133.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 139.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 125.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 135.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 122.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 140.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 131.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 137.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 123.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 127.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 120.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 114.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 122.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 109.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 112.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 231.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 222.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 228.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 219.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 227.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 240.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 220.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 224.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 239.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 216.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 202.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 192.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 224.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 213.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 190.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 184.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 198.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 194.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 224.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 216.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 186.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 183.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 193.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 184.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 198.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 189.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 188.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 204.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 181.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 194.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 209.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 187.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 176.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 167.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 201.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 188.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 164.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 155.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 179.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 167.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 204.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 189.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 166.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 158.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 178.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 169.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 183.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 174.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 173.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 188.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 166.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 179.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 194.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 171.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 162.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 150.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 185.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 171.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 149.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 139.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 169.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 155.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 192.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 178.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 155.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 144.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 210.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 203.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 211.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 316.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 327.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 311.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 181.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 171.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 196.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 182.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 190.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 183.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 171.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 165.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 236.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 227.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 232.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 223.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 231.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 245.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 224.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 229.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 244.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 221.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 206.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 196.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 229.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 221.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 195.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 188.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 202.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 198.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 226.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 220.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 191.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 186.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 197.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 188.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 203.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 193.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 193.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 208.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 185.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 199.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 213.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 191.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 181.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 171.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 205.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 192.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 170.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 159.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 183.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 171.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 208.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 194.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 172.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 182.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 173.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 187.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 178.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 178.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 193.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 170.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 183.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 199.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 175.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 166.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 154.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 189.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 175.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 155.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 142.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 173.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 160.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 196.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 182.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 148.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 100.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 88.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 106.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 94.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 100.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 88.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 106.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 94.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 111.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 99.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 105.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 122.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 104.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 132.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 113.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 107.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 103.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 122.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 110.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 106.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 106.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 113.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 99.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 95.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 116.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 104.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 101.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 86.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 89.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 123.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 111.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 105.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 90.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 105.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 93.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 119.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 113.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 108.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 99.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 129.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 111.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 111.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 100.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 98.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 117.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 104.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 97.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 90.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 114.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 102.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 99.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 84.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 97.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 85.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 106.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 94.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 120.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 108.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 121.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 89.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 101.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 107.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 97.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 128.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 105.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 116.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 110.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 101.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 109.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 100.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 109.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 99.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 96.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 118.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 104.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 115.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 102.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 108.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 95.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 100.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 88.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 112.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 100.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 97.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 82.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 95.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 83.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 104.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 92.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 118.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 107.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 119.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 106.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 101.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 86.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 87.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 119.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 130.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 118.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 112.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 101.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 117.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 104.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 109.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 96.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 89.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 113.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 101.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 98.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 84.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 84.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 105.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 105.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 93.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 119.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 107.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 88.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 106.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 115.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 102.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 133.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 104.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 114.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 101.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 123.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 104.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 92.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 114.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 100.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 100.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 88.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 106.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 122.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 102.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 130.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 101.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 110.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 97.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 120.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 108.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 96.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 99.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 87.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 110.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 95.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 95.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 83.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 114.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 103.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 120.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 100.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 109.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 96.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 128.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 99.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 108.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 95.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 94.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 97.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 85.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 108.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 93.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 93.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 81.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 102.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 111.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 98.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 130.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 114.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 101.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 110.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 97.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 108.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 95.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 98.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 87.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 108.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 94.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 114.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 114.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 104.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 122.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 111.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 112.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 102.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 119.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 108.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 111.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 100.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 117.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 106.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 111.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 101.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 118.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 107.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 238.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 92.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 96.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 83.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 230.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 83.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 90.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 89.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 81.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 229.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 216.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 85.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 72.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 91.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 78.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 86.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 75.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 82.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 69.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 123.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 137.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 120.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 111.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 121.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 110.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 107.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 127.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 115.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 125.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 110.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 120.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 128.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 116.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 124.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 112.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 107.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 105.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 93.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 113.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 134.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 122.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 129.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 117.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 110.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 96.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 110.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 97.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 126.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 104.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 134.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 124.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 118.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 107.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 124.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 122.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 107.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 117.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 104.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 94.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 125.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 113.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 89.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 89.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 111.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 131.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 119.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 125.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 113.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 93.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 105.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 93.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 106.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 113.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 102.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 133.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 110.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 116.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 105.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 105.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 105.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 123.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 120.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 116.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 102.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 92.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 124.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 112.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 118.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 106.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 103.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 88.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 98.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 121.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 109.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 130.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 118.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 123.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 106.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 91.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 103.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 90.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 127.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 135.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 124.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 107.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 110.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 117.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 104.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 105.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 93.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 126.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 114.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 89.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 100.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 88.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 123.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 131.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 119.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 125.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 93.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 104.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 92.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 133.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 111.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 123.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 106.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 133.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 121.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 131.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 107.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 119.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 102.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 130.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 129.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 105.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 116.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 128.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 116.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 131.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 107.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 118.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 102.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 132.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 104.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 89.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 110.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 94.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 105.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 88.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 110.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 94.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 115.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 100.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 121.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 105.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 126.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 110.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 132.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 116.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 120.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 105.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 126.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 111.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 117.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 101.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 123.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 107.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 126.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 110.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 132.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 116.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 120.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 104.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 126.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 110.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 117.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 101.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 123.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 107.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 137.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 122.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 143.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 127.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 132.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 117.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 137.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 122.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 128.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 113.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 134.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 118.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 107.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 96.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 113.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 102.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 107.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 96.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 113.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 101.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 111.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 98.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 116.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 103.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 110.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 99.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 115.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 104.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 160.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 161.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 174.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 147.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 155.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 160.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 163.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 174.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 146.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 156.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 147.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 159.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 171.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 136.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 153.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 147.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 161.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 171.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 135.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 155.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 145.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 134.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 159.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 147.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 171.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 159.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 137.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 124.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 151.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 141.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 141.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 134.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 158.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 170.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 158.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 132.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 123.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 148.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 140.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 141.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 145.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 155.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 128.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 138.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 143.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 147.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 159.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 129.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 141.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 128.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 142.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 152.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 117.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 137.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 130.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 145.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 155.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 119.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 139.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 126.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 115.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 143.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 127.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 154.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 139.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 116.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:24 104.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 132.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 121.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 128.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 117.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 145.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 130.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 157.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 143.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 117.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 105.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 135.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 124.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 132.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 135.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 119.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 129.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 135.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 140.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 151.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 122.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 133.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 119.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 133.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 143.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 108.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 127.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 122.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 137.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 147.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 111.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 131.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 116.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 105.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 133.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 118.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 144.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 130.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 107.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 95.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 123.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 113.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 120.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 109.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 137.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 122.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 149.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 135.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 109.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 98.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 127.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 116.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 171.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 166.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 178.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 155.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 159.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 171.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 166.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 177.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 154.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 159.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 158.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 163.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 175.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 144.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 158.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 158.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 164.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 174.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:24 143.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 158.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 155.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 145.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 162.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 148.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 174.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 160.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 144.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 130.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 154.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 152.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 147.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 160.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 148.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 173.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 161.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 139.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 133.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 151.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 142.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 152.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 148.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 159.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 136.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 142.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 154.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 151.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 162.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 137.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 144.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 139.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 145.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 155.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 125.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 141.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 149.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 158.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 126.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 136.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 124.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 145.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 128.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 157.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 140.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 123.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 110.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 135.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 122.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 138.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 126.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 148.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 131.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 160.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 144.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 124.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 112.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 137.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 125.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 143.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 138.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 150.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 127.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 132.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 146.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 143.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 154.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 129.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 136.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 130.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 136.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 116.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 131.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 133.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 140.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 151.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 119.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 134.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 127.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 115.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 136.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 119.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 148.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 131.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 114.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 101.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 126.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 113.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 130.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 118.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 140.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 123.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 152.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 136.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 117.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 104.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 130.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 117.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 147.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 144.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 133.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 124.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 108.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 99.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 152.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 185.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 163.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 197.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 145.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 178.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 154.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 185.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 166.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 198.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 147.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 177.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 146.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 146.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 151.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 184.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 160.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 194.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 144.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 177.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 153.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 184.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 163.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 195.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 146.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 177.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 151.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 138.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 173.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 170.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 163.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 148.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 187.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 181.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 142.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 139.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 146.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 133.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 167.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 164.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 152.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 172.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 171.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 164.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 151.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 188.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 182.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 142.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 139.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 146.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 134.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 166.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 164.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 140.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 158.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 151.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 169.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 133.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 150.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 144.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 161.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 155.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 173.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 136.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 153.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:24 118.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 119.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 139.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 157.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 148.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 167.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 133.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 150.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 142.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 160.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 152.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 171.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 136.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 152.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 139.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 125.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 154.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 144.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 151.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 135.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 167.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 155.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:24 122.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 112.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 134.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 120.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 148.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 137.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 142.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 129.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 154.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 146.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 154.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 169.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 158.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 121.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 114.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 136.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 123.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 148.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 135.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 146.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 146.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 158.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 128.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 139.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 139.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 149.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 151.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 162.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 132.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 141.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 106.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 108.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 134.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 145.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 143.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 155.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 128.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 138.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 138.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 148.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 148.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 159.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 131.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 141.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 134.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 120.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 142.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 132.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 145.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 131.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 154.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 108.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 100.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 129.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 115.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 136.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 125.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 137.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 124.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 145.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 135.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 150.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 135.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 158.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 146.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 110.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 102.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 132.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 119.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 139.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 128.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 137.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 145.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 133.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 133.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 135.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 142.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 131.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 136.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 123.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 144.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 130.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 139.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 127.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 131.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 121.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 155.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 188.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 167.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 200.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 148.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 181.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 158.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 188.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 169.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 201.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 150.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 180.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 154.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 154.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 154.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 187.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 163.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 197.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 147.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 180.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 156.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 187.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 166.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 198.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 149.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 180.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 153.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 139.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 178.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 160.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 166.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 149.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 190.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 171.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 149.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 138.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 148.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 133.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 172.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 154.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 154.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 140.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 175.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 161.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 167.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 151.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 191.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 172.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 149.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 138.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 149.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 134.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 169.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 154.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 143.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 161.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 155.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 173.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 136.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 154.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 147.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 164.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 159.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 177.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 139.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 156.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 125.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 127.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 142.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 160.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 151.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 170.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 135.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 153.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 145.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 163.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 156.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 174.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 139.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 155.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 141.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 126.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 157.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 139.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 153.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 136.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 170.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 149.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 129.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 117.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 136.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 120.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 151.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 132.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 144.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 129.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 157.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 140.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 157.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 139.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 172.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 152.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 129.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 118.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 139.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 123.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 151.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 133.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 138.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 149.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 149.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 161.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 131.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 142.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 153.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 154.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 165.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 135.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 144.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 114.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 116.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 137.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 148.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 146.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 158.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 130.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 141.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 141.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 151.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 151.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 162.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 134.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 144.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 136.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 121.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 144.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 125.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 148.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 131.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 157.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 136.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 116.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 104.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 131.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 139.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 119.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 140.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 124.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 147.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 129.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 152.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 135.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 161.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 140.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 118.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 107.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 134.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 118.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 141.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 122.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 166.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 163.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 151.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 133.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 138.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 122.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 109.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 166.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 206.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 185.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 225.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 159.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 200.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 167.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 207.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 188.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 226.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 161.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 200.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 149.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 164.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:30 153.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 164.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 163.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 204.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 178.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 220.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 158.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 198.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 165.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 205.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 181.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 221.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 159.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 198.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 164.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 150.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 194.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 191.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 180.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 169.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 211.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 208.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 159.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 162.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 143.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 157.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 157.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 144.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 187.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 184.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 164.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 152.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 190.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 191.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 181.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 171.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 208.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 208.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:33 164.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 158.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 147.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 158.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 158.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 183.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 184.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 151.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 175.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 170.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 195.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 145.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 168.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 155.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 178.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 175.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 199.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 148.78 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 171.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 132.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 131.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 137.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 134.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 149.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 172.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 163.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 188.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 143.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 166.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 153.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 176.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 168.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 192.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 147.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 170.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 149.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 135.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 169.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 159.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 164.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 154.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 186.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 176.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 142.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 137.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 127.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 126.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 142.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 129.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 163.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 152.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 153.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 139.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 170.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 163.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 169.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 158.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 188.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 180.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 146.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:24 138.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:24 130.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 128.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 146.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 133.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 163.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 156.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 145.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 159.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 165.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 180.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 139.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 153.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 149.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 163.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 170.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 184.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 143.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 156.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 125.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 117.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 129.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 119.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 143.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 157.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 158.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 173.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 137.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 151.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 147.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 161.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 163.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 178.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 142.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 155.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 143.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 130.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 155.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 144.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 158.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 148.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 171.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 164.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 134.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 121.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 119.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 111.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 137.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 123.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 148.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 138.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 147.76 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 134.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 159.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 148.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 163.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 153.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 176.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 168.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:24 138.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 124.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 123.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 114.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 141.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 128.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 151.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 141.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 154.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 163.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 150.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 145.74 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 152.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 152.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 160.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 149.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 150.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 139.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 162.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 147.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 169.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 160.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 139.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 146.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 146.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 137.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 169.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 210.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 188.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 229.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 162.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 203.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 170.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 210.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 191.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 230.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 164.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 203.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 156.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 172.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 160.81 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 171.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 166.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 208.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 182.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 223.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 201.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 168.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 208.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 184.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 224.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 162.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 201.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 166.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 150.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 198.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 183.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 183.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 168.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 215.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 198.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 167.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 169.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 151.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 156.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 159.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 144.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 191.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 176.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 167.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 153.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 193.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 184.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 184.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 171.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 212.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 200.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 171.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 165.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:30 154.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 156.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 160.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 146.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 185.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 176.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 154.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 178.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 174.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 198.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 148.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 171.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 158.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 181.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 178.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 202.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 151.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 174.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 139.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 139.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 143.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 142.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 152.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 176.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 167.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 191.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 170.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 156.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 179.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 171.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 195.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 150.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 173.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 152.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 135.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 172.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 157.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 167.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 153.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 190.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 172.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 148.92 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 144.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:26 131.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 130.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 145.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 129.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 165.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 150.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 156.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 139.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 173.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 159.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 172.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 158.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 191.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 175.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 152.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 145.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 136.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 132.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 149.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 133.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 166.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 151.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 148.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 163.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 168.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 183.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 156.44 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 152.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 167.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 173.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 188.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 146.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 159.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 132.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 124.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 136.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 127.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.36 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 161.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 161.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 176.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 140.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 154.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 150.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 165.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 166.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 181.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 144.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 158.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 146.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 130.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 158.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 141.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 161.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 145.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 174.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 159.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 140.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 129.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:24 124.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 114.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 139.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 123.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 151.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 134.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 150.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 134.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 162.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 145.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 166.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 152.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 179.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 164.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 144.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 132.04 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 128.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 117.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 143.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 128.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 154.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 138.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 198.54 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 195.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 182.48 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 156.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 160.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 147.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 138.87 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 141.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 129.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 188.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 247.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 180.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 238.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 190.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 246.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 182.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 237.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 197.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 197.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 185.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 243.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 207.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 266.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 178.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 236.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 187.29 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 242.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 209.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 266.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 179.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 235.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 187.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 173.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 229.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 230.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 211.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 195.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 257.62 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 253.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 194.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 191.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 178.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 165.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 220.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 222.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 188.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 174.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 220.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 229.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 211.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 197.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 246.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 253.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:24 183.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 191.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 178.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 166.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 211.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 221.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 171.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 203.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 163.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 195.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 176.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 210.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 167.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 201.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 155.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 159.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 167.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 200.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 189.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 222.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 160.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 192.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 173.02 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 206.90 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 195.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 230.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 165.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 199.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 169.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 154.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 198.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 187.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 191.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 176.79 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 222.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 210.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 160.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 149.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 160.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 147.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 188.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 179.39 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 174.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 159.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 198.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 193.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 197.25 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 183.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 222.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 217.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 158.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 152.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 165.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 152.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 188.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 185.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 163.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 185.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 155.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 177.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 169.46 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 191.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 160.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 182.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:24 137.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 141.31 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 160.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 182.49 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 181.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 205.24 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 153.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 175.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 166.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 188.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 188.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 211.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 158.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 180.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 161.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 179.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 169.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 184.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 169.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 203.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 193.01 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 141.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 131.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 152.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 139.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 170.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 161.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 167.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 153.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 184.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 174.93 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 190.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 176.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 208.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 199.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 144.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 135.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 158.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 145.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 174.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 166.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 175.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 168.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 176.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 171.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 182.18 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 166.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 175.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 158.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 188.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 169.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 183.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 170.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 164.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 156.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 192.17 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 250.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 183.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 241.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 193.82 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 249.63 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 185.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 240.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 204.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 205.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 188.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 246.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 210.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 269.60 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 181.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 239.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 190.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 246.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 212.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 269.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 183.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 238.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 189.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 162.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 232.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 218.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 213.12 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 183.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 258.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 240.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 202.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 188.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 180.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 154.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 223.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 210.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 190.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 164.51 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 223.50 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 215.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 213.64 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 186.85 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 248.21 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 238.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 191.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 183.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 181.26 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 156.68 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 213.52 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 207.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 174.30 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 207.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 165.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 198.22 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 179.56 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 213.55 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 170.84 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 204.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 163.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 166.66 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 170.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 203.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 193.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 226.43 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 163.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 196.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 175.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 210.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 198.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 233.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 168.83 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 202.00 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 171.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 144.35 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 200.80 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 183.72 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 194.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 165.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 224.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 206.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 168.14 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 152.69 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 162.88 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 136.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 191.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 175.34 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 176.91 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 150.59 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 201.19 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 185.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 199.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 172.57 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 225.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 208.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 165.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 152.97 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 167.70 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 142.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 190.96 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 176.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 166.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 189.16 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 158.32 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 180.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 172.67 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 194.65 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 163.95 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 185.38 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 145.20 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 149.06 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 163.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 185.71 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 185.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 208.73 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 156.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 178.13 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 169.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 191.33 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 191.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 214.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 161.94 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 183.11 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 163.99 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 138.05 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 182.77 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 165.98 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 186.75 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 159.45 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 206.10 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 188.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 149.03 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 134.58 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 155.07 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 130.37 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 173.09 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 157.61 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 170.23 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 144.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 187.42 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 170.86 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 193.28 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 166.27 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 211.53 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 194.08 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 152.15 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 138.40 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 160.89 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 136.47 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 177.41 KB
fmhaSm100fKernel_QkvBfloat16OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 162.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 100.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 88.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 104.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 92.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 99.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:25 107.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:29 94.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:31 135.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:25 123.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:31 109.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:30 96.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 88.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 103.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:29 111.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:25 98.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:26 139.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:33 126.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:29 112.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:24 100.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 92.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 110.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 99.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 114.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 103.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 130.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 134.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 113.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 117.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 105.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 121.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 119.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 105.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 109.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 93.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 116.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 104.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 100.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 87.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 120.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 113.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 98.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 120.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 108.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 126.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 112.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 94.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 91.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 128.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 104.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 131.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 114.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 119.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 102.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 96.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 118.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 100.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 116.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 89.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 113.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 102.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 116.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 103.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 88.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 95.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 83.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 105.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 109.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 93.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 117.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 106.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 121.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 103.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 91.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 99.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 87.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 126.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 95.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 130.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 106.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 100.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 109.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 100.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 94.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 114.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 98.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 101.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 87.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 112.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 100.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 114.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 100.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 87.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 93.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 81.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 104.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 106.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 91.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 104.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 119.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 102.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 85.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 128.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 114.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 112.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 97.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 131.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 115.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 119.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 102.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 96.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 119.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 116.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 100.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 103.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 88.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 113.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 101.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 100.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 88.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 94.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 82.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 93.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 117.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 105.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 103.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 91.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 98.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 86.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 135.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 107.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 116.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 138.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 126.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 106.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 114.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 99.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 126.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 105.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 90.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 114.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 102.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 98.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 86.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 132.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 103.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 112.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 96.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 135.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 123.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 101.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 110.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 101.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 101.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 86.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 110.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 98.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 93.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 81.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 116.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 104.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 101.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 109.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 94.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 133.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 121.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 99.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 108.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 93.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 121.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 100.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 98.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 84.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 107.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 95.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 91.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 79.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 114.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 103.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 132.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 102.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 112.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 96.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 135.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 123.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 101.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 110.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 95.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 123.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 115.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 101.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 99.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 88.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 108.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 98.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 92.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 83.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 115.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 115.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 104.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 119.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 108.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 112.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 102.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 116.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 106.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 111.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 100.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 115.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 104.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 112.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 101.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 105.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 215.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 86.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 94.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 98.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 82.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 204.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 84.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 91.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 89.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 80.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 204.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 190.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 84.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 92.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 79.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 89.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 76.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 82.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 138.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 123.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 106.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 141.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 129.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 123.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 110.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 129.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 120.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 111.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 105.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 128.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 116.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 125.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 109.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 127.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 130.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 118.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 125.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 97.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 91.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 117.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 132.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 120.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 130.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 117.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 112.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 95.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 135.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 109.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 138.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 124.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 105.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 126.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 108.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 126.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 111.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 105.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 124.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 93.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 127.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 115.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 94.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 87.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 124.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 129.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 117.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 126.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 96.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 90.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 117.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 137.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 111.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 122.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 119.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 124.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 105.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 114.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 98.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 124.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 109.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 119.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 109.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 108.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 91.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 126.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 114.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 105.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 84.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 95.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 127.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 115.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 123.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 95.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 101.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 88.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 136.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 109.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 119.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 101.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 139.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 121.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 126.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 116.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 109.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 100.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 110.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 92.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 127.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 116.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 106.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 98.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 86.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 112.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 112.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 97.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 129.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 117.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 125.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 112.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 108.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 96.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 90.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 144.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 122.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 142.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 129.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 141.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 108.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 118.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 139.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 139.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 105.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 116.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 98.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 137.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 141.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 107.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 118.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 100.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 141.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 100.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 88.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 104.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 92.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 99.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:33 107.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:26 94.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:26 135.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:28 123.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:32 109.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:26 96.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 88.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 103.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:29 111.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:32 98.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:28 139.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:33 126.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:33 113.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:31 100.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 92.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 96.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 84.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 99.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 87.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 97.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 85.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 100.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 88.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 108.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 110.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 98.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 135.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 118.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 130.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 128.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 140.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 131.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 116.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 135.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 118.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 106.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 100.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 125.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 138.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 126.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 91.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 113.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 121.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 109.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 129.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 142.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 130.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 106.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 94.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 116.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 120.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 117.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 129.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 110.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 124.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 134.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 113.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 99.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 109.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 112.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 129.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 113.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 111.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 106.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 100.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 94.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 131.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 97.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 85.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 94.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 114.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 110.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 103.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 98.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 122.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 136.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 123.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 100.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 114.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 126.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 107.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 121.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 131.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 112.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 121.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 106.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 126.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 99.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 110.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 108.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 97.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 91.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 115.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 128.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 94.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 91.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 111.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 100.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 119.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 105.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 133.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 120.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 97.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 85.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 95.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 108.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 109.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 122.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 120.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 132.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 97.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 110.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 118.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 128.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 100.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 110.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 98.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 115.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 128.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 95.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 83.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 91.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 100.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 96.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 134.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 98.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 86.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 121.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 107.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 124.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 110.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 112.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 101.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 116.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 103.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 109.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 98.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 112.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 100.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 110.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 98.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 113.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 101.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 147.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 102.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 111.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 119.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 98.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 135.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 99.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 134.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 121.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 100.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 86.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 109.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 95.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 108.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 94.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 95.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 83.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 135.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 129.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 140.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 122.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 137.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 133.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 144.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 124.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 126.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 135.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 130.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 139.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 113.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 134.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 122.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 118.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 129.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 143.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 97.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 135.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 124.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 121.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 133.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 146.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 133.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 99.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 124.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 128.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 134.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 126.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 139.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 120.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 119.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 124.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 134.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 107.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 118.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 126.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 111.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 122.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 136.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 121.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 102.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 90.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 127.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 114.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 126.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 112.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 140.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 127.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 93.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 102.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 119.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 131.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 112.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 123.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 136.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 116.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 113.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 116.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 126.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 101.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 116.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 131.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 104.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 115.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 123.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 108.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 99.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 133.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 118.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 87.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 110.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 124.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 114.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 111.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 123.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 137.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 123.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 102.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 91.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 126.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 132.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 113.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 130.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 137.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 114.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 128.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 102.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 118.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 122.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 133.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 124.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 109.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 119.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 134.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 100.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 95.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 125.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 115.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 112.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 124.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 138.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 124.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 92.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 97.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 84.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 100.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 87.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 97.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 85.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 100.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 88.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 124.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 131.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 119.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 119.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 122.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 129.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 118.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 109.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 96.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 115.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 103.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 113.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 101.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 104.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta128PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 94.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 105.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 92.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 108.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 95.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 106.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 93.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 109.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 96.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 135.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 130.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 121.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 121.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 107.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 137.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 131.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 124.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 121.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 115.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 105.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 122.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 110.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 109.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 96.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 125.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 121.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 113.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 112.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 107.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 98.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 111.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 101.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 118.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 106.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 91.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 120.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 108.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 95.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 138.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 148.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 134.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:26 118.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 121.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 136.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 145.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 133.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 124.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 110.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 158.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 119.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 118.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 116.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 105.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 104.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 118.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 107.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 136.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 122.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 139.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 125.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 124.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 111.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 127.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 114.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 107.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 110.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 131.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 132.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 137.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 126.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 113.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 138.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 123.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 111.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 112.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 114.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 101.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 116.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 115.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 106.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 96.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 124.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 113.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 100.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 171.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 162.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 172.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 163.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 167.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 179.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 160.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 168.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 183.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 161.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 171.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 155.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 190.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 177.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 156.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 144.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 173.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 157.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 195.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 180.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 157.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 144.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 155.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 146.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 160.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 151.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 151.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 164.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 144.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 156.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 171.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 149.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 156.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 139.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 175.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 161.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 140.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 127.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 161.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 145.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 182.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 166.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 145.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 133.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 148.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 140.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 154.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 145.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 144.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 158.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 137.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 150.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 165.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 143.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 149.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 133.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 169.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 154.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 134.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 121.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 155.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 139.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 175.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 161.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 139.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 127.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 156.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 150.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 141.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 153.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 163.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 147.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 143.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 127.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 154.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 138.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 135.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 122.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 131.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvDenseDynamicTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 124.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 175.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 166.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 178.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 169.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 171.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 184.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 164.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 175.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 187.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 167.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 175.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 159.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 195.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 180.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 164.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 147.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 177.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 162.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 199.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 184.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 166.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 147.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 159.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 150.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 165.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 155.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 155.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 169.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 148.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 161.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 176.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 153.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 160.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 144.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 180.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 163.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 149.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 130.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 165.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 149.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 186.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 170.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 154.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 136.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 153.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 144.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 159.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 149.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 149.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 162.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 142.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 155.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 170.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 147.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 153.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 137.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 173.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 156.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 143.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 124.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 159.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 143.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 179.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 163.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 148.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 130.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 94.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 82.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 98.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 86.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 94.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:26 110.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:32 87.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:31 138.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:26 116.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:25 112.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:28 89.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 82.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 98.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:26 110.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:33 91.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:32 139.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:30 120.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:28 113.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:27 93.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 86.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 105.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 93.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 109.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 97.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 114.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 108.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 100.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 123.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 104.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 100.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 117.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 103.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 94.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 91.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 111.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 98.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 83.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 98.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 86.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 95.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 115.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 102.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 87.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 90.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 114.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 99.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 110.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 100.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 97.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 95.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 92.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 86.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 108.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 95.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 96.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 81.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 93.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 82.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 109.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 97.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 112.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 99.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 99.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 84.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 85.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 108.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 94.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 119.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 103.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 96.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 106.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 96.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 93.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 109.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 90.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 84.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 106.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 94.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 97.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 94.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 79.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 91.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 79.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 95.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 100.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 88.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 110.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 98.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 98.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 82.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 95.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 83.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 110.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 95.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 121.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 114.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 108.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 98.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 95.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 101.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 105.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 91.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 97.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 85.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 107.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 95.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 98.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 95.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 80.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 80.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 96.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 90.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 111.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 99.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 99.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 84.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 96.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 85.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 121.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 102.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 112.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 99.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 112.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 101.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 113.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 117.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 107.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 93.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 100.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 88.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 110.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 98.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 96.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 110.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 98.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 95.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 124.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 97.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 108.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 93.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 91.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 95.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 83.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 106.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 94.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 91.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 80.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 107.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 95.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 105.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 92.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 95.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 106.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 91.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 89.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 93.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 81.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 104.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 92.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 89.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 78.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 106.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 94.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 119.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 98.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 107.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 94.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 124.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 110.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 96.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 93.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 103.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 90.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 94.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 104.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 94.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 90.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 81.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 95.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 109.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 98.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 113.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 102.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 106.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 95.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 110.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 99.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 104.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 94.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 108.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 98.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 105.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 95.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 109.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 99.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 209.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 82.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 90.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 95.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 81.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 201.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 81.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 88.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 87.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 80.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 200.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 185.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 83.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 70.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 89.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 76.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 85.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 74.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 81.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 68.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 124.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 108.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 130.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 113.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 124.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 115.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 125.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 112.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 101.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 89.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 119.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 107.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 111.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 126.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 114.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 108.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 93.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 122.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 115.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 100.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 128.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 99.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 118.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 106.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 113.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 90.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 122.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 109.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 116.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 101.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 86.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 97.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 85.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 116.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 104.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 106.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 111.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 121.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 109.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 105.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 89.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 102.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 89.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 98.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 126.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 106.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 111.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 105.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 101.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 111.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 98.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 100.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 88.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 120.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 108.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 114.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 102.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 84.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 95.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 83.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 92.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 122.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 119.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 103.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 87.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 100.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 128.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 113.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 118.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 107.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 113.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 89.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 109.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 115.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 101.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 86.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 96.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 84.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 116.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 104.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 123.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 111.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 120.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 89.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 101.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 88.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 130.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 107.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 119.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 103.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 130.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 127.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 103.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 115.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 98.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 127.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 126.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 101.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 113.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 96.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 128.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 103.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 115.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 130.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 116.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:33 110.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:31 87.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:27 138.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:29 116.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:30 112.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:27 89.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:25 111.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:26 91.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:32 139.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:31 120.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:28 113.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:33 93.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 98.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 86.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 102.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 90.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 98.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 86.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 102.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 90.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 103.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 91.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 107.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 94.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 102.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 92.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 107.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 96.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 131.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 132.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 142.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 117.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 126.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 137.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 136.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 145.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 122.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 129.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 120.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 129.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 139.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 105.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 124.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 126.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 133.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 111.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 128.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 118.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 106.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 131.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 114.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 141.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 126.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 104.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 92.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 120.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 108.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 122.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 112.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 133.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 117.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 144.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 129.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 111.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 97.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 121.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 111.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 121.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 123.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 133.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 106.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 117.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 124.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 127.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 138.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 109.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 110.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 121.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 130.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 95.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 115.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 125.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 135.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 97.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 119.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 108.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 97.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 121.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 105.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 131.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 117.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 94.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 82.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 111.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 111.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 100.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 125.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 109.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 136.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 122.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 96.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 84.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 114.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 103.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 116.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 119.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 129.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 101.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 113.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 119.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 123.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 135.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 104.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 117.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 105.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 117.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 126.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 90.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 111.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 108.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 121.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 131.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 93.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 116.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 103.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 92.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 117.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 101.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 127.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 113.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 89.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 77.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 107.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 96.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 106.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 95.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 121.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 106.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 133.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 119.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 91.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 79.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 111.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 100.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 142.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 135.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 146.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 129.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 148.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 139.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 149.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 129.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 132.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 131.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 133.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 142.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 113.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 127.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 137.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 136.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 146.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 118.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 131.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 129.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 116.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 133.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 143.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 128.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 112.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 98.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 123.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 110.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 132.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 122.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 136.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 118.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 147.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 130.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 119.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 103.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 124.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 112.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 132.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 126.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 137.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 114.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 120.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 135.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 130.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 142.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 116.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 124.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 124.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 133.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 103.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 118.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 124.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 128.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 138.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 105.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 122.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 119.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 106.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 124.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 106.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 135.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 118.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 101.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 88.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 113.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 101.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 121.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 109.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 128.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 111.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 139.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 123.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 103.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 90.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 117.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 105.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 127.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 122.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 133.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 109.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 116.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 130.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 127.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 138.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 112.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 120.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 116.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 120.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 129.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 98.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 114.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 119.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 125.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 135.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 100.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 119.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 113.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 101.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 120.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 102.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 131.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 114.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 96.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 83.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 97.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 116.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 104.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 124.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 107.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 135.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 120.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 99.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 86.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 114.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk320HV256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 101.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 115.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 103.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 98.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 91.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 93.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 99.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 87.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 126.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 139.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 137.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 151.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 119.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 132.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 129.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 141.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 141.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 154.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 122.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 134.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 107.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 109.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 125.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 138.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 134.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 148.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 118.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 131.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 128.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 140.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 138.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 151.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 121.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 133.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 124.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 111.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 138.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 124.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 137.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 122.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 151.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 135.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 102.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 119.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 106.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 132.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 118.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 127.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 138.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 126.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 140.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 125.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 153.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 138.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 113.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 103.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 122.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 109.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 133.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 120.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 121.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 128.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 132.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 140.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 114.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 121.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 125.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 131.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 136.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 144.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 118.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 124.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 97.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 99.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 120.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 127.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 129.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 137.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 114.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 120.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 124.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 130.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 133.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 141.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 117.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 123.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 119.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 107.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 126.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 113.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 132.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 117.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 139.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 124.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 102.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 91.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 114.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 101.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 120.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 108.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 123.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 110.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 129.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 116.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 136.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 121.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 143.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 128.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 104.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 93.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 117.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 104.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 123.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 110.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 119.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 123.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 130.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 135.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 112.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 116.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 122.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 127.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 134.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 139.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 115.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 119.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 92.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 94.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 118.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 122.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 127.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 132.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 111.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 121.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 126.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 131.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 136.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 114.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 118.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 117.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 104.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 121.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 109.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 129.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 114.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 134.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 120.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 97.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 86.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 112.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 99.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 115.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 103.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 121.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 108.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 125.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 112.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 134.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 119.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 139.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 124.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 99.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 89.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 115.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 119.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 106.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 113.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 120.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 109.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 108.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 110.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 117.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 107.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 112.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 120.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 105.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 115.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 103.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 108.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 96.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 129.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 143.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 140.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 155.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 122.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 135.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 132.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 145.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 144.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 157.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 125.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 137.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 115.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 116.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 128.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 141.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 137.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 152.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 121.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 134.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 131.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 143.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 141.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 154.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 124.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 136.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 127.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 112.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 141.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 125.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 140.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 122.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 154.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 136.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 108.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 121.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 106.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 135.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 119.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 130.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 115.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 141.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 126.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 143.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 126.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 156.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 138.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 120.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 108.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 109.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 135.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 120.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 124.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 131.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 135.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 143.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 117.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 124.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 128.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 135.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 139.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 147.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 121.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 127.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 104.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 107.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 123.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 130.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 133.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 116.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 123.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 127.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 133.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 137.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 144.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 120.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 126.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 122.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 107.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 129.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 113.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 135.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 118.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 142.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 124.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 109.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 97.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 117.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 102.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 123.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 108.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 125.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 111.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 132.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 117.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 139.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 122.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 146.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 128.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 111.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 99.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 120.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 105.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 126.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 111.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 122.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 127.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 133.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 138.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 119.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 126.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 130.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 137.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 143.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 118.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 122.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 99.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 102.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 121.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 125.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 130.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 135.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 114.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 118.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 124.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 129.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 134.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 140.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 117.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 121.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 120.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 105.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 124.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 109.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 132.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 115.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 137.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 120.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 104.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 92.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 114.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 99.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 118.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 103.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 123.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 109.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 127.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 113.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 136.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 119.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 142.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 124.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 107.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 95.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 118.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 103.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 121.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 106.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 113.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 119.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 90.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 92.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 97.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 135.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 151.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 153.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 171.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 129.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 145.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 139.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 153.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 158.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 173.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 133.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 108.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 112.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 111.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 114.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 133.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 149.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 147.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 164.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 127.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 144.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 137.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 151.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 152.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 167.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 131.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 145.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 133.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 120.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 150.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 136.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 150.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 135.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 167.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 155.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 116.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 118.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 103.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 106.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 127.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 113.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 143.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 130.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 137.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 124.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 151.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 137.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 154.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 139.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 169.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 157.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 119.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 119.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 105.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 108.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 130.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 117.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 144.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 130.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 130.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 138.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 148.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 157.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 123.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 132.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 134.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 142.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 153.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 162.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 128.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 135.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 101.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 97.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:26 104.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 99.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 127.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 136.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 142.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 151.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 122.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 130.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 132.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 140.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 147.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 156.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 126.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 134.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 127.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 114.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 136.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 123.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 144.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 129.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 153.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 139.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 108.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 104.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 95.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 92.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 121.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 108.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 129.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 116.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 132.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 119.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 140.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 127.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 149.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 135.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 158.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 146.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 111.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 105.92 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:26 98.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 94.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 125.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 112.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 133.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 120.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 127.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 132.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 145.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 152.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 120.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 126.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 131.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 137.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 150.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 157.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 125.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 130.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 96.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 90.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 99.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 93.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 124.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 130.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 139.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 145.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 119.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 125.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 129.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 135.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 144.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 151.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 123.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 129.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 125.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 111.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 130.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 117.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 141.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 126.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 147.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 133.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:33 103.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 96.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 90.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 84.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 118.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 105.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 123.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 111.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 129.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 116.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 135.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 122.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 146.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 132.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 153.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 139.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 105.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 98.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 93.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 87.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 123.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 109.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 128.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 115.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 122.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 130.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 118.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 106.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 110.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 119.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 127.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 116.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 119.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 106.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 129.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 114.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:32 117.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 120.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:30 100.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 105.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 115.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 104.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 138.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 155.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 157.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 174.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 132.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 148.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 142.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 156.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 161.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 177.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 136.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 149.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 115.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 119.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 117.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 122.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 136.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 153.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 150.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 168.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 130.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 147.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 140.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 154.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 155.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 170.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 134.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 148.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 136.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 120.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 152.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 136.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 152.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 135.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 170.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 154.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:26 123.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 125.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:26 107.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 110.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 129.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 114.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 145.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 130.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 140.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 124.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 154.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 138.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 157.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 140.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 172.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 156.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:33 125.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 126.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:25 110.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 113.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 133.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 118.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 146.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 131.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 133.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 141.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 151.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 161.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 126.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 135.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 137.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 145.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 156.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 166.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 131.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 138.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 108.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 105.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 110.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 107.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 131.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 139.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 145.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 154.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 125.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 133.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 135.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 143.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 150.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 159.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 129.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 137.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 130.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 115.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 139.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 123.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 147.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 129.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 156.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 138.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:26 114.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 111.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 101.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 95.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 124.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 108.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 132.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 116.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 135.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 120.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 143.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 127.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 152.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 135.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 161.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 146.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:30 117.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 113.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 103.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 97.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 128.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 113.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 135.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 121.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 130.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 136.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 148.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 155.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 123.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 129.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 134.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 140.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 153.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 161.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 128.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 133.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 103.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:25 98.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 106.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 100.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 128.01 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 134.18 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 142.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 149.30 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 122.44 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 128.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 132.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 138.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 147.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 154.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 126.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 132.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 127.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 112.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 133.51 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 118.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 144.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 127.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 151.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 133.61 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:31 109.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 103.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:28 96.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 88.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 121.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 106.21 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 126.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 111.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 132.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 117.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 138.12 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 122.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 149.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 132.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 156.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 139.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Persistent2CtaKeepsAbForGen.cubin 22-Jul-2026 16:29 112.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 106.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128Static2CtaKeepsAbForGen.cubin 22-Jul-2026 16:27 99.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 91.13 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 125.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 110.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 130.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 115.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 134.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 124.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 100.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 106.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 94.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 152.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 175.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 144.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 167.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 158.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 176.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 150.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 167.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 128.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 126.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 149.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 172.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 171.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 194.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 142.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 165.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 155.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 173.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 177.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 196.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 147.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 165.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 151.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 136.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 174.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 159.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 173.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 158.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 198.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 182.60 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:25 134.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 122.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 141.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 128.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 165.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 151.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 156.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 141.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 174.96 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 160.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 179.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 164.58 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 199.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 183.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:31 130.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 121.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 146.91 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 134.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 165.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 152.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 145.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 158.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 137.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 149.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 151.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 163.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 143.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 155.07 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 108.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 109.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 142.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 154.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 163.76 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 176.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 135.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 147.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 148.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 160.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 170.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 183.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 141.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 152.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 143.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 129.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 156.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 141.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 166.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 151.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 179.39 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 164.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 115.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 102.87 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 134.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 121.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 146.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 133.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 149.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 135.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 161.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 147.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 172.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 158.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 185.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 170.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 116.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 103.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 140.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 127.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 151.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 139.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 141.95 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 150.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 133.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 142.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 148.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 157.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 139.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 148.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 99.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 102.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 138.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 147.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 160.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 169.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 131.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 145.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 154.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 167.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 176.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 137.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 146.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 140.29 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 125.54 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 148.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 134.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 162.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 147.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 172.08 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 157.14 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:28 106.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:33 93.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 131.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 117.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 139.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 126.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 146.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 132.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 155.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 140.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 169.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 154.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 179.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 164.36 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 108.97 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 96.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 137.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 124.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 145.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 132.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 138.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 131.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 124.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 133.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 144.68 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 128.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 140.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 121.86 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 150.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 132.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 133.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 118.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 128.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvDenseStaticTokenSparseP1VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 118.23 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 156.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 178.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 147.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 170.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 161.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 180.00 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 152.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 170.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:30 135.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 134.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 152.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 175.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 174.27 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 198.24 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 145.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 168.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 158.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 176.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 180.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 199.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 150.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 168.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 153.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 132.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 177.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 147.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 176.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 153.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 200.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 170.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:29 141.81 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 129.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 144.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 124.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 167.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 140.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 159.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 137.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 177.34 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 151.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 182.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 159.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 202.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 174.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:26 137.80 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:29 131.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 149.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 129.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 167.37 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 143.42 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 148.79 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 161.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 140.45 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 152.77 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 155.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 167.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 146.46 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 158.15 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 116.17 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 117.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 145.47 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 158.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 166.99 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 180.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 138.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 150.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 151.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 163.72 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 173.93 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 187.10 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 144.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 155.88 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 146.55 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 125.09 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 159.02 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 131.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 169.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 146.31 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 182.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 153.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:27 122.90 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:26 108.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 137.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 117.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 149.59 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 123.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 152.53 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 131.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 164.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 138.20 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 175.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 153.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 188.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 161.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:32 124.05 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:28 110.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 143.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 123.19 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 154.41 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 129.85 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 145.03 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 154.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 136.70 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 145.62 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 151.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 160.65 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 142.83 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 151.63 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 107.04 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:31 109.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 141.84 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 150.78 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 163.22 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 173.25 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 134.43 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 143.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 148.11 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 157.33 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 170.28 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 180.35 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 140.69 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 149.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 142.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 121.32 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 151.71 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 125.49 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 165.56 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 142.52 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 175.06 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 147.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:30 113.50 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:27 99.73 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 133.75 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 113.64 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 142.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 117.66 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 149.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 127.74 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 157.82 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 132.16 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 172.48 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 149.38 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 182.26 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 154.89 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentKeepsAbForGen.cubin 22-Jul-2026 16:33 116.40 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticKeepsAbForGen.cubin 22-Jul-2026 16:32 102.57 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 139.94 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 119.67 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 147.98 KB
fmhaSm100fKernel_QkvE4m3OBfloat16HQk576HV512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 123.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 101.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 89.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 105.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 93.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 101.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:33 109.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:32 96.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:32 137.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:33 124.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:32 111.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:29 98.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 89.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 105.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:31 113.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:26 99.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:28 141.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:28 128.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:28 115.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:32 102.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 93.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 112.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 100.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 116.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 104.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 133.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 110.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 119.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 136.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 114.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 124.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 124.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 108.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 117.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 123.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 122.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 109.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 106.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 125.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 113.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 96.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 118.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 105.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 94.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 89.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 127.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 116.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 100.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 109.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 124.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 112.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 97.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 105.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 93.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 130.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 105.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 114.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 98.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 133.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 110.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 120.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 116.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 98.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 118.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 122.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 115.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 103.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 115.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 91.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 85.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 124.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 119.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 107.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 107.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 94.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 88.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 128.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 132.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 100.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 120.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 110.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 95.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 115.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 105.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 99.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 105.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 90.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 113.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 101.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 101.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 90.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 82.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 123.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 110.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 94.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 117.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 105.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 118.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 106.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 106.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 93.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 86.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 130.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 115.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 114.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 98.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 134.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 110.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 117.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 97.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 117.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 107.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 102.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 121.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 106.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 91.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 114.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 102.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 114.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 92.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 83.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 118.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 119.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 107.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 107.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 94.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 99.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 87.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 137.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 108.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 117.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 102.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 140.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 129.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 107.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 100.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 127.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 112.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 109.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 94.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 115.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 103.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 99.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 87.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 125.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 134.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 104.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 113.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 97.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 137.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 126.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 103.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 111.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 96.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 124.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 104.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 89.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 110.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 99.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 95.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 83.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 133.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 102.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 111.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 95.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 135.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 100.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 109.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 94.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 102.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 87.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 108.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 97.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 93.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 81.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 109.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 135.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 104.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 113.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 97.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 137.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 126.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 102.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 111.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 96.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 125.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 108.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 103.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 91.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 109.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 99.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 93.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 84.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 117.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 105.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 109.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 114.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 103.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 118.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 107.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 112.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 101.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 116.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 105.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 113.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 102.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 117.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 140.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 143.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 119.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 129.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 132.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 122.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 131.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 127.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 110.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 132.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 132.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 119.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 112.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 92.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 134.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 122.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 133.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 121.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 129.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 109.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 137.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 140.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 125.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 129.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 109.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 128.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 123.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 129.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 129.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 116.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 88.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 131.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 130.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 118.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 125.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 136.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 108.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 139.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 123.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 127.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 127.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 115.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 108.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 96.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 86.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 117.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 98.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 90.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 138.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 120.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 141.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 114.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 125.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 129.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 123.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 113.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 129.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 129.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 116.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 87.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 132.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 130.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 118.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 91.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 147.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 113.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 123.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 105.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 147.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 137.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 144.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 109.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 119.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 101.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 144.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 134.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 142.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 107.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 117.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 99.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 143.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 133.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 144.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 109.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 119.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 101.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 146.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 134.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 101.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:24 109.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:24 96.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:24 137.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 124.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:24 111.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:24 98.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 89.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 105.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:23 113.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 100.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:23 141.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:23 128.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:24 115.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:23 102.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 93.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 97.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 85.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 100.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 88.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 98.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 86.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 101.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 88.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 109.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 96.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 111.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 99.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 129.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 126.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 137.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 132.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 129.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 142.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 121.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 118.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 127.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 137.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 119.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 107.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 130.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 116.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 139.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 127.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 90.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 122.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 134.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 144.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 131.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 93.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 118.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 107.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 131.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 123.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 136.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 131.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 112.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 101.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 95.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 122.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 132.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 84.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 115.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 103.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 127.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 137.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 87.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 123.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 98.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 108.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 109.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 97.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 92.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 129.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 116.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 81.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 112.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 100.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 96.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 134.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 121.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 84.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 129.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 122.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 134.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 113.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 115.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 124.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 109.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 129.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 111.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 98.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 93.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 129.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 81.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 114.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 110.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 135.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 85.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 122.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 125.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 114.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 101.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 116.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 104.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 110.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 113.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 101.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 111.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 99.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 114.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 102.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 137.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 131.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 142.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 139.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 134.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 146.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 126.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 127.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 124.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 137.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 114.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 122.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 127.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 132.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 141.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 125.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 135.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 119.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 134.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 144.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 132.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 95.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 136.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 121.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 138.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 123.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 149.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 136.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 117.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 98.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 127.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 136.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 117.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 133.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 141.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 121.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 117.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 121.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 131.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 136.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 127.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 112.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 127.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 137.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 89.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 115.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 131.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 142.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 129.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 92.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 127.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 130.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 125.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 138.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 118.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 118.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 114.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 118.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 128.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 118.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 123.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 133.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 124.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 109.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 123.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 134.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 121.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 86.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 125.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 115.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 112.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 139.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 89.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 134.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 132.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 126.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 139.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 130.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 113.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 135.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 125.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 110.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 100.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 134.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 121.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 105.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 87.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 126.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 113.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 104.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 139.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 127.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 90.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 93.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 109.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 96.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 107.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 94.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 110.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 97.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 135.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 131.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 122.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 117.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 126.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 138.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 132.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 125.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 103.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 123.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 111.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 106.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 114.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 94.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 126.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 114.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 119.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 114.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 107.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 90.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 121.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 109.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 93.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 137.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 123.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 140.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 126.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 125.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 112.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 128.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 115.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 120.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 107.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 110.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 131.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 133.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 138.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 125.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 131.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 140.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 133.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 130.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 115.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 95.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 172.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 163.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 174.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 165.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 168.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 183.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 161.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 170.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 186.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 162.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 179.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 165.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 195.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 181.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 161.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 150.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 182.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 168.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 198.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 183.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 161.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 150.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 156.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 147.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 162.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 153.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 152.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 168.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 145.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 158.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 175.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 150.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 164.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 150.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 179.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 164.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 145.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 133.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 170.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 155.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 185.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 170.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 150.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 138.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 150.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 141.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 156.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 147.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 146.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 161.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 139.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 152.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 168.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 145.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 157.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 143.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 173.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 157.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 139.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 126.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 163.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 149.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 179.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 164.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 144.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 176.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 167.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 180.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 171.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 173.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 187.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 165.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 176.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 191.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 168.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 184.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 169.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 199.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 185.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 170.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 153.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 186.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 172.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 202.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 188.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 172.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 153.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 161.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 152.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 166.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 157.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 157.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 173.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 150.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 163.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 179.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 155.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 169.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 154.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 184.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 168.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 155.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 136.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 174.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 159.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 190.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 175.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 159.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 142.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 155.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 145.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 160.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 151.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 151.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 166.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 143.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 157.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 173.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 149.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 162.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 147.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 177.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 162.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 148.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 130.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 167.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 153.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 184.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 169.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 153.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 136.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 82.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 104.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 86.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 103.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:24 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 89.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:24 129.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:23 117.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:23 102.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:23 91.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 82.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 104.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:23 104.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 92.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 133.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:24 121.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 106.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:23 94.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 86.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 113.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 93.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 114.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 97.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 125.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 110.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 118.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 104.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 119.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 98.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 119.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 86.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 87.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 102.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 89.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 91.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 116.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.06 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 100.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 96.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 88.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 116.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 95.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 83.33 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 95.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 83.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 113.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 99.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 86.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 87.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.03 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 95.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 103.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 113.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 108.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.22 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 97.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 94.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 111.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 112.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 86.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 114.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 94.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 96.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 81.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 93.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 81.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 102.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 90.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 116.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 98.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 85.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 85.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 107.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 122.81 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 99.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 95.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 87.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 115.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 95.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 82.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 93.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 82.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 100.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 91.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 117.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 99.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 118.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 86.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.83 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 86.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 103.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 113.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 100.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 128.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 102.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 114.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 99.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 101.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 90.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 114.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 102.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 98.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 86.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 102.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 99.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 109.43 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 96.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 98.40 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 110.07 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 95.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 95.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 97.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 85.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 109.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 97.75 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 93.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 82.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 97.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 107.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 93.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 124.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 96.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 107.90 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 93.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 114.98 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 93.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 95.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 83.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 107.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 95.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 91.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 79.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 98.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 99.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 109.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 95.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 126.69 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 112.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 98.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 109.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 94.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 94.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 96.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 87.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 108.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 97.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 92.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 83.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 118.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 120.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 102.56 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 115.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 95.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 99.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 113.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.39 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 115.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 98.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 114.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 95.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 116.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 126.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 121.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 116.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 132.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 114.30 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 125.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 108.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.34 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 108.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 130.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 112.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.35 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 91.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 123.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.42 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 112.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 114.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 128.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 116.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 110.19 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 94.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 95.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 105.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.89 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 113.50 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 129.80 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 121.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 114.47 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 115.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 100.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.49 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 92.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 127.66 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 109.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 119.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 88.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 87.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 129.25 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 111.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.60 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 107.36 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.86 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 90.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.51 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.59 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.97 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 99.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.26 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.74 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.94 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 89.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 126.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 108.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.28 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.32 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 102.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 87.10 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.52 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 84.96 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 107.17 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 94.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 127.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 109.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 122.46 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.31 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.82 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 90.45 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 88.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 124.18 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.72 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.67 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 113.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 130.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.99 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.23 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.62 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 104.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.13 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.79 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.61 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.16 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.00 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 104.78 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 91.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 127.57 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 109.84 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.77 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.68 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 88.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.01 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 86.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.93 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.92 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.54 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 95.73 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 129.04 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 111.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.08 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 111.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.02 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 92.11 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 90.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 132.20 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 108.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 121.09 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 104.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 133.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.29 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 129.38 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 104.41 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 116.88 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 100.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.27 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.58 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 127.85 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 102.24 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 114.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 97.76 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 128.71 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.05 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 129.95 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 104.48 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 116.70 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 99.87 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 132.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.64 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:24 100.55 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:24 89.21 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:24 129.37 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:24 118.12 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:23 102.65 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:23 91.15 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:23 104.44 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 92.91 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:24 133.14 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:23 121.63 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:23 106.53 KB
fmhaSm100fKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:23 94.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 105.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 98.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 105.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:24 120.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 108.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:23 148.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:24 136.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:23 122.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:24 110.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 94.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 109.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 124.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 111.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:23 152.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:24 140.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:24 126.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:24 114.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 98.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 116.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 105.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 120.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 109.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 131.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 134.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 123.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 100.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 92.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 122.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 110.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 102.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 87.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 126.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 114.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 91.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 128.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 132.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 96.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 116.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 88.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 119.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 107.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 88.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 83.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 115.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 107.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 123.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 111.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 87.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 127.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 95.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 86.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 117.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 99.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 98.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 86.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 81.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 90.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 121.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 110.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 89.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 84.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 129.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 132.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 87.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 118.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 107.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 99.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 87.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 93.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 82.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 105.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 91.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 122.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 110.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 105.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 90.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 86.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 135.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 107.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 116.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 100.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 138.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 127.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 105.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 114.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 99.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 104.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 89.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 111.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 100.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 97.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 85.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 118.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 133.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 102.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 112.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 96.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 135.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 124.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 101.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 110.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 95.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 99.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 85.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 107.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 96.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 92.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 81.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 131.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 100.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 109.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 93.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 134.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 99.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 108.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 93.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 97.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 83.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 105.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 94.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 90.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 79.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 114.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 133.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 102.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 112.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 95.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 136.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 101.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 109.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 94.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 98.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 87.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 105.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 96.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 91.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 82.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 121.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 110.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 125.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 114.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 118.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 107.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 122.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 111.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 116.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 106.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 120.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 117.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 106.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 121.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 110.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 138.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 123.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 142.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 109.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 129.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 125.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 136.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 124.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 96.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 91.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 125.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 113.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 138.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 125.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 126.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 136.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 119.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 118.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 139.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 113.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 121.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 127.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 116.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 111.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 104.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 133.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 121.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 93.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 86.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 123.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 134.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 123.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 95.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 90.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 134.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 137.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 122.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 125.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 119.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 89.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 131.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 120.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 103.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 84.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 133.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 121.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 94.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 88.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 136.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 139.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 116.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 126.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 91.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 133.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 121.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 105.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 93.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 86.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 123.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 135.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 123.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 121.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 95.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 89.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 145.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 111.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 122.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 104.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 139.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 129.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 142.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 107.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 117.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 100.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 136.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 126.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 141.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 105.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 115.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 97.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 135.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 143.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 107.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 117.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 99.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 138.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 105.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:24 120.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 108.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:23 149.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:23 136.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:24 122.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:23 110.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 94.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 109.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:23 124.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 112.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:23 152.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:24 140.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 126.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:24 114.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 101.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 89.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 104.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 92.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 90.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 105.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 93.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 113.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 100.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 116.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 103.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 141.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 125.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 136.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 118.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 143.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 128.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 140.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 131.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 125.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 125.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 135.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 123.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 111.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 121.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 90.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 100.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 126.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 114.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 139.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 125.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 93.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 134.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 129.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 137.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 122.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 134.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 100.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 129.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 116.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 105.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 99.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 127.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 84.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 119.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 107.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 132.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 100.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 87.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 109.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 98.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 131.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 134.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 119.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 131.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 121.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 126.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 113.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 109.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 102.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 96.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 93.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 81.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 90.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 116.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 104.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 101.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 129.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 97.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 84.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 132.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 127.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 136.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 113.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 118.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 118.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 128.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 115.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 102.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 82.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 91.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 118.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 105.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 129.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 98.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 85.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 126.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 112.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 129.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 118.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 105.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 121.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 108.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 115.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 103.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 105.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 115.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 103.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 118.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 106.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 148.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 129.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 140.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 124.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 151.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 132.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 145.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 125.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 135.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 133.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 130.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 140.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 139.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 127.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 123.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 127.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 140.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 108.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 96.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 104.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 140.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 129.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 126.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 131.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 145.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 130.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 141.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 122.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 134.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 145.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 126.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 139.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 129.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 127.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 134.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 131.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 119.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 116.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 101.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 90.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 113.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 132.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 123.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 119.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 124.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 138.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 124.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 93.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 117.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 138.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 119.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 131.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 142.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 123.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 136.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 126.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 111.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 124.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 121.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 131.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 128.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 116.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 113.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 130.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 87.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 94.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 116.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 121.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 135.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 121.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 90.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 139.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 132.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 143.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 137.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 117.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 128.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 111.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 125.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 122.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 133.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 130.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 114.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 129.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 87.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 95.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 130.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 117.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 121.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 134.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 121.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 91.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 111.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 99.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 110.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 97.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 113.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 100.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 98.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 114.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 101.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 131.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 132.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 140.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 135.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 126.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 107.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 143.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 136.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 129.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 120.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 128.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 115.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 95.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 131.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 118.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 113.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 98.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 115.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 124.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 118.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 111.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 90.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 126.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 122.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 113.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 94.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 141.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 127.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 144.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 130.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 116.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:24 132.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 119.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:23 125.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:23 111.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:24 114.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 139.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 140.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 121.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 143.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 129.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 125.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 144.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 130.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 125.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 112.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 127.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 130.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 112.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 131.29 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 134.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 116.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 126.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.89 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 96.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 116.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 112.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 99.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 171.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 162.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 172.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 163.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 167.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 180.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 160.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 168.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 183.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 161.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 167.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 150.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 184.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 169.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 155.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 143.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 170.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 151.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 188.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 173.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 156.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 144.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 155.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 146.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 160.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 151.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 151.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 165.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 144.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 156.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 172.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 149.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 152.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 133.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 169.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 154.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 140.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 126.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 157.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 139.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 175.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 160.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 145.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 132.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 148.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 139.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 154.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 145.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 144.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 158.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 137.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 150.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 165.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 143.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 146.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 127.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 163.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 147.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 133.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 151.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 134.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 170.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 154.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 139.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 126.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 175.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 166.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 178.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 169.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 171.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 184.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 164.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 175.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 188.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 167.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 171.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 154.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 191.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 174.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 164.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 147.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 174.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 154.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 195.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 177.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 166.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 147.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 159.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 150.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 165.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 155.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 155.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 169.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 148.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 161.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 176.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 153.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 156.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 137.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 176.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 158.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 149.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 130.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 161.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 143.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 182.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 165.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 153.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 136.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 153.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 144.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 159.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 149.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 149.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 163.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 142.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 155.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 170.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 147.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 150.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 131.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 169.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 151.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 142.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 123.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 156.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 137.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 176.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 158.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 147.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 130.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 94.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 82.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 98.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 86.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 94.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 104.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:24 92.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:24 132.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:23 121.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:23 106.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:24 94.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 82.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 98.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:23 107.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:23 96.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:24 136.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:24 124.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:23 110.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:23 98.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 86.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 105.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 93.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 109.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 103.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 100.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 123.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 108.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 99.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 114.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 116.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 90.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 111.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 98.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 83.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 86.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 95.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 115.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 102.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 120.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 101.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 86.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 101.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 89.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 114.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 95.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 98.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 94.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 100.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 99.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 91.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 98.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 86.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 108.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 95.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 95.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 80.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 93.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 81.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 96.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 90.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 112.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 99.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 115.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 99.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 83.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 97.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 85.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 93.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 119.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 104.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 96.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 106.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 96.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 92.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 109.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 110.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 90.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 84.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 106.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 94.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 93.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 78.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 91.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 79.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 88.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 110.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 98.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 82.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 83.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 109.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.95 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 95.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 114.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 108.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 99.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 108.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 111.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 102.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 112.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 98.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 103.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 97.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 85.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 95.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 109.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 94.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 80.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 80.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 108.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 95.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 89.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 111.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 114.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 98.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 83.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 95.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 84.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 102.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 112.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 98.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 127.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 112.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 101.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 112.74 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 97.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 105.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 93.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 99.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 87.87 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 109.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 97.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 96.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 84.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 98.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 119.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:22 98.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 107.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 94.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 97.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 108.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 93.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 102.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 90.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 95.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 83.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 105.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 93.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 91.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 79.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 107.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 95.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 95.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 105.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 92.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 123.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 108.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 94.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 106.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 91.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 89.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:24 93.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 81.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:22 103.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 91.19 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 89.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 77.63 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 94.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 119.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 97.83 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 107.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 94.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 125.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 110.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 96.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 108.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:24 93.11 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 101.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 90.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 93.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 85.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 103.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 93.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:23 90.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:23 81.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 107.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 95.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 113.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 106.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 95.92 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 110.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 99.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 104.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 108.66 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 98.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 105.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 95.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 109.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 124.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 119.45 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 114.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 130.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 117.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 121.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 122.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 113.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 94.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 125.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 112.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 107.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 88.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 89.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 118.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 106.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 110.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 98.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 126.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 114.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 124.50 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 112.67 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 107.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 91.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 105.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 93.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 122.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 104.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 115.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 128.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 120.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 104.17 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 112.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 103.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.57 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 104.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 99.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 118.58 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 107.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 118.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 106.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 111.07 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 99.49 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 102.34 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 90.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 122.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:23 109.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 115.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 103.56 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 100.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 85.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 96.99 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 85.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 115.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 103.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.60 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:24 123.90 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 120.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 108.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 104.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 89.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.51 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 89.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 120.59 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 101.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 113.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 110.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 126.77 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 106.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 117.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 113.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.88 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 111.37 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 111.40 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 97.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 117.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 105.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 116.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 105.18 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 109.53 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 97.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 100.00 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 88.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:23 120.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:24 108.28 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:24 113.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 101.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 99.25 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 84.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 94.80 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 83.10 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:23 114.31 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:24 102.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:22 104.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:23 92.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 122.36 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 109.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 118.04 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 87.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 99.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 86.85 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 122.69 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.02 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.26 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 128.64 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.46 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.96 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 114.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 104.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.47 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.09 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.35 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.44 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.05 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.43 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.01 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 99.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.61 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 89.79 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 122.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 109.98 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.78 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.62 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 101.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 85.84 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 84.68 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 116.08 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 103.93 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 105.75 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 93.94 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 123.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 111.54 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.65 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.38 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 104.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 89.22 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 100.81 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 88.55 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 130.52 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 107.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 119.42 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 102.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 128.72 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 119.13 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.70 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 103.03 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 115.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 98.41 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.86 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 126.16 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 100.73 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.91 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 96.24 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 124.32 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 114.76 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 128.27 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.97 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 115.15 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 98.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 128.14 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 116.48 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:30 104.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:31 92.39 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:32 133.06 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:33 121.30 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:33 106.20 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:28 94.33 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:27 108.12 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:32 96.23 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:30 136.71 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:31 124.82 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:26 110.21 KB
fmhaSm100fKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:31 98.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 100.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 88.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 104.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 92.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 100.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:33 107.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:28 94.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:32 135.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:30 123.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:29 109.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:32 96.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 88.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 104.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:29 111.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:27 98.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:33 139.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:25 126.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:33 113.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:33 100.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 92.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 99.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 114.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 103.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 131.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 108.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 115.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 101.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 134.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 112.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 123.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 123.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 100.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 122.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 120.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 107.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 104.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 109.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 93.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 117.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 105.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 103.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 91.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 120.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 109.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 98.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 121.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 109.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 126.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 94.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 91.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 128.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 104.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 114.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 97.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 132.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 116.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 100.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 115.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 89.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 114.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 102.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 100.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 95.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 83.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 117.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 109.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 93.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 117.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 106.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 121.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 91.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 87.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 127.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 101.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 110.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 95.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 130.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 106.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 113.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 101.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 117.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 104.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 114.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 102.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 112.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 101.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 114.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 101.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 87.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 93.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 81.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 116.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 116.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 104.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 89.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 97.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 85.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 129.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 115.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 101.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 119.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 88.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 113.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 101.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 100.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 94.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 82.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 105.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 93.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 117.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 105.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 120.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 91.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 97.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 86.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 135.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 107.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 116.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 100.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 138.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 106.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 114.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 99.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 116.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 105.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 90.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 114.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 102.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 97.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 85.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 108.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 96.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 135.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 124.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 101.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 110.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 95.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 122.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 101.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 86.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 110.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 93.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 81.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 116.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 131.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 100.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 109.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 93.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 134.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 123.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 99.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 108.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 93.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 121.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 112.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 101.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 98.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 107.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 95.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 91.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 79.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 133.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 102.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 112.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 95.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 136.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 101.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 109.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 94.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 123.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 113.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 101.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 99.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 88.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 108.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 98.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 91.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 83.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 104.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 115.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 105.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 119.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 109.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 112.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 102.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 116.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 106.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 111.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 101.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 115.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 105.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 112.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 101.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 116.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 105.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 138.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 123.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 121.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 106.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 142.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 128.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 124.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 130.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 111.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 120.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 111.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 129.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 115.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 125.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 109.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 97.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 131.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 118.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 125.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 109.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 97.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 91.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 117.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 132.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 120.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 130.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 112.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 99.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 94.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 136.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 118.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 101.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 139.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 124.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 121.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 108.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 100.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 126.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 111.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 121.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 122.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 110.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 93.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 128.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 116.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 121.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 106.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 94.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 87.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 125.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 113.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 129.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 117.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 126.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 109.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 97.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 90.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 134.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 106.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 137.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 122.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 126.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 105.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 114.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 125.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 109.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 119.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 91.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 126.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 114.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 105.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 92.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 96.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 84.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 123.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 111.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 95.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 127.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 116.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 123.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 108.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 95.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 88.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 136.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 101.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 139.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 113.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 124.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 105.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 128.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 109.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 100.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 121.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 123.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 128.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 116.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 120.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 106.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 94.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 86.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 97.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 129.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 117.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 112.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 97.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 102.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 90.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 145.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 122.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 104.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 140.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 130.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 142.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 107.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 118.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 100.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 137.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 127.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 141.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 105.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 97.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 136.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 126.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 143.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 107.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 117.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 100.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 140.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 127.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 100.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:27 107.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:32 94.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:28 135.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:26 123.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:27 109.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:28 96.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 88.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 104.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:32 111.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:33 98.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:29 139.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:30 126.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:29 113.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:33 100.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 92.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 96.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 85.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 99.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 87.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 97.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 85.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 100.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 88.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 108.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 95.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 98.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 136.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 130.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 128.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 140.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 122.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 131.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 125.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 135.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 109.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 119.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 118.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 106.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 101.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 139.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 126.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 100.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 121.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 109.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 129.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 114.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 143.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 130.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 95.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 121.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 117.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 111.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 134.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 124.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 109.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 129.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 103.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 113.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 111.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 100.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 95.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 132.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 98.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 86.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 94.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 114.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 103.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 98.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 122.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 137.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 101.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 89.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 110.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 114.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 126.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 121.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 119.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 131.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 97.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 116.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 126.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 100.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 110.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 108.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 104.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 97.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 91.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 115.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 115.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 95.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 83.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 91.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 111.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 100.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 95.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 119.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 134.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 121.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 98.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 86.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 95.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 115.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 123.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 120.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 133.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 113.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 123.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 110.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 118.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 128.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 101.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 110.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 104.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 98.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 99.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 129.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 96.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 91.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 113.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 101.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 96.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 120.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 134.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 120.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 99.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 87.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 96.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 121.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 107.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 124.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 110.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 113.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 101.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 116.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 104.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 110.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 113.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 101.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 110.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 98.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 113.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 101.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 135.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 140.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 123.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 122.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 138.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 132.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 145.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 125.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 125.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 123.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 126.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 135.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 130.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 140.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 114.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 134.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 122.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 118.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 115.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 144.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 128.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 104.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 135.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 124.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 121.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 112.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 133.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 147.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 134.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 100.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 124.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 128.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 134.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 132.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 126.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 139.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 119.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 119.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 116.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 120.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 129.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 105.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 114.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 124.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 134.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 108.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 126.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 112.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 102.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 122.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 105.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 137.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 122.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 91.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 97.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 127.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 118.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 114.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 126.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 112.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 141.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 94.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 131.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 129.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 123.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 136.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 113.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 126.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 111.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 131.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 105.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 114.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 123.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 109.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 99.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 102.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 133.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 119.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 100.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 110.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 124.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 111.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 123.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 138.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 124.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 103.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 91.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 114.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 120.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 132.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.54 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 130.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 137.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 117.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 128.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 103.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 112.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 122.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 133.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 116.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 125.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 110.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 100.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 134.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 101.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 89.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 95.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 125.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 112.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 124.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 139.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 92.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 112.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 100.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 105.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 93.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 108.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 95.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 106.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 93.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 109.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 96.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 135.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 130.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 122.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 138.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 131.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 124.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 122.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 109.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 112.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 103.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 123.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 118.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 111.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 105.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 110.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 97.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 126.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 122.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 113.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 109.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 107.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 102.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 118.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 106.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 101.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 121.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 109.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 95.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 137.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 123.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 139.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 125.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 124.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 112.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 127.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 114.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 120.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 107.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 122.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 110.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 131.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 133.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 138.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 127.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 139.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 109.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 123.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 112.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 126.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 113.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 118.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 115.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 104.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 107.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 108.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 97.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 125.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 114.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 100.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 171.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 162.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 172.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 163.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 167.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 180.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 160.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 168.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 183.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 161.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 171.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 156.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 191.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 177.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 156.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 144.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 174.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 158.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 196.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 180.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 157.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 144.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 155.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 146.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 160.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 151.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 151.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 165.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 144.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 156.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 172.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 149.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 156.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 140.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 176.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 161.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 140.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 127.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 161.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 145.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 182.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 167.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 145.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 133.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 148.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 139.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 154.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 145.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 144.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 158.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 137.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 150.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 165.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 143.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 150.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 133.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 170.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 154.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 134.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 155.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 139.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 176.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 161.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 139.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 127.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 175.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 166.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 179.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 169.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 171.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 184.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 164.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 175.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 188.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 167.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 175.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 160.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 196.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 180.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 164.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 147.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 178.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 162.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 200.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 184.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 166.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 147.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 159.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 150.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 165.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 155.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 156.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 169.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 148.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 161.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 176.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 153.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 160.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 144.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 181.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 163.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 149.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 130.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 165.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 149.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 187.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 170.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 154.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 136.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 153.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 144.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 159.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 149.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 149.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 163.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 142.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 155.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 170.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 147.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 153.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 138.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 174.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 156.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 143.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 124.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 160.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 143.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 180.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 164.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 148.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 130.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 95.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 83.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 99.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 87.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 95.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:32 111.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:31 87.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:27 138.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:29 116.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:32 113.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:27 89.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 83.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 99.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:33 111.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:30 91.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:28 140.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:33 120.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:33 113.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:29 93.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 87.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 106.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 94.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 110.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 98.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 100.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 123.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 114.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 103.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 108.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 95.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 90.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 112.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 99.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 116.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 99.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 84.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 98.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 86.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 100.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 95.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 116.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 103.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 121.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 108.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 102.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 101.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 90.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 114.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 99.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 110.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 95.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 120.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 108.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 97.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 105.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 86.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 109.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 96.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 112.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 96.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 81.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 93.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 81.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 97.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 113.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 100.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 84.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 97.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 85.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 97.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 93.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 97.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 103.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 96.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 92.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 100.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 110.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 99.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 91.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 84.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 107.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 95.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 97.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 94.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 79.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 91.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 79.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 108.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 95.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 88.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 111.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 99.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 114.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 102.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 98.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 83.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 95.40 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 83.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 115.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 99.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.87 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 105.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 95.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 121.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.01 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 98.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 98.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 94.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 102.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 101.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 91.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 97.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 85.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 108.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 96.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 110.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 95.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 80.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 92.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 80.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 109.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 89.97 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 112.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 100.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 103.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 84.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 96.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 84.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 122.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 112.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 98.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 127.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 101.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 112.85 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 97.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 117.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.43 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 94.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 100.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 88.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 110.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 98.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 96.29 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 84.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 98.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 119.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 98.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 107.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 94.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 124.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 109.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 97.00 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 108.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 93.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 103.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 91.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 95.58 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 83.65 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 106.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 94.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 91.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 80.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 96.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 95.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 105.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 92.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 123.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 94.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 106.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 91.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 113.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 89.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 93.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 81.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 103.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 92.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 89.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 77.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.63 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.71 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 97.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 107.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 94.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 125.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 110.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 96.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 108.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 93.22 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 91.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 94.14 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 85.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 104.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 94.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 90.18 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 81.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 95.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 110.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 99.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 114.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 103.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 107.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 96.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 111.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 100.80 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 105.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 95.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 109.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 99.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 106.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 96.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 110.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 100.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 124.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 108.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 119.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.50 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 131.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 124.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 117.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 108.37 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 115.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.76 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 122.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 109.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 102.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 94.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 126.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 113.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 121.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 108.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 89.19 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 101.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 89.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 119.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 98.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 127.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 114.91 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.04 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 93.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 122.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 115.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 111.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 100.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 128.28 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 120.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 112.89 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.68 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 99.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 118.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 90.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 123.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 110.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 101.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 86.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 97.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 85.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 116.86 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 106.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 94.82 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 124.78 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 112.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 121.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 109.56 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 89.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 101.77 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 89.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 102.09 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.05 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 110.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 97.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 126.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 106.49 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 117.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 111.51 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.60 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 117.15 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 116.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 110.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 100.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 88.26 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 121.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 109.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 100.17 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 84.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 83.38 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.32 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.31 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 104.44 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 123.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 110.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 88.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.46 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 122.67 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 104.16 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 115.12 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 112.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 100.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 128.75 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 108.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 120.07 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 114.81 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.45 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 103.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.47 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.55 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 99.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 119.03 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.61 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 118.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.99 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.23 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.59 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.88 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 89.96 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 122.98 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 110.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.64 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 86.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 96.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 84.83 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 117.21 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.06 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 105.92 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 94.11 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 124.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 112.42 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 120.94 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.39 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.93 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 89.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 88.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 130.62 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 107.35 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 119.53 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 102.74 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 130.10 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.57 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 127.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 103.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 115.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 98.52 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.24 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 116.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 126.27 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 100.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 113.02 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 96.36 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 125.70 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.33 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 128.25 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 102.95 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 115.13 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 98.34 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.41 KB
fmhaSm100fKernel_QkvE4m3OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 116.79 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:33 111.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:27 87.66 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:27 138.90 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:30 116.69 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:33 113.30 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:26 89.72 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:30 111.84 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:31 91.48 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:30 140.73 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:27 120.20 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:31 114.08 KB
fmhaSm100fKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:28 93.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 106.37 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 94.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 111.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 100.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 105.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 94.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 110.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 100.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 116.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 105.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 122.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 111.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 139.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.89 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 133.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 126.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 115.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 149.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 127.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 138.54 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.43 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 119.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 131.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 130.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 115.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 136.93 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 125.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 135.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 118.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 124.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 128.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 116.10 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 137.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 101.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.67 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 136.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 123.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 127.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 133.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 121.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 140.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 126.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 117.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 134.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 115.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 121.17 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.10 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 144.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 120.22 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 131.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 124.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 126.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 111.08 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 132.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 127.53 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.43 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 111.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.60 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 110.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 127.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 113.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 110.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 106.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 92.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 130.22 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.78 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 120.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.08 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 127.57 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 116.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 133.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 113.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 110.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 96.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 130.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 120.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 141.40 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 114.47 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 125.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 120.63 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 106.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 123.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.35 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 107.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 101.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 106.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 105.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 119.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 107.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 92.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 100.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 87.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 126.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 114.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 124.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 112.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 127.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 109.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 104.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 91.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 132.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 111.10 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 121.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.74 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 142.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 115.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 126.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 122.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 109.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.08 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 113.89 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 123.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 112.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 110.40 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 119.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 107.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 122.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.93 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 106.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 93.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 87.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 127.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 98.57 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 124.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 112.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 127.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 109.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 104.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 91.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 142.82 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 121.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 130.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 114.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 151.22 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 134.51 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 120.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 128.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 113.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 138.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 113.54 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 120.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 105.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 130.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 114.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 113.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 99.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 133.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 121.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 137.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 114.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 122.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 106.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 146.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 129.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 112.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 121.04 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 106.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 120.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 112.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 97.28 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 121.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 106.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 104.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 91.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 127.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 116.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 134.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 117.67 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 101.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 142.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 126.02 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 107.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 116.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 100.87 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 130.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 116.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.74 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 107.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 92.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 116.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 101.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 98.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 87.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 123.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 113.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 135.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 109.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 118.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 102.51 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 143.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 108.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 116.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 101.65 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 131.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 116.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.02 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 106.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 92.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 115.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 101.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 98.47 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 124.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 113.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 127.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 116.23 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 132.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 121.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 121.99 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 110.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 126.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 115.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 118.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 107.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 123.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 112.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 118.53 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 107.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 123.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 113.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 147.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 127.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 138.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.50 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.99 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 155.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 131.79 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 143.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 135.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 123.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 139.00 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 126.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 134.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 122.37 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 119.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 142.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 129.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 139.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 122.90 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 134.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 121.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 128.35 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 110.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 139.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 127.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 140.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 118.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 141.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 132.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 116.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 144.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 132.69 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 144.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 130.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 109.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 141.83 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 130.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 127.87 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 150.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 124.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 135.96 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 130.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 116.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.74 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.68 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 127.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 111.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 137.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 122.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 132.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 120.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 116.03 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 128.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 115.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 120.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 102.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 133.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 122.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 132.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 115.67 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.27 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 109.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 95.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 135.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 123.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 125.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 106.78 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 138.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 127.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 137.38 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 123.57 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 119.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 138.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 114.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 107.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 147.06 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 130.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 126.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 130.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.34 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 122.26 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.00 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 134.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.53 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 127.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.89 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 112.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 114.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.09 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 130.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 118.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 126.61 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.76 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 91.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 132.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 120.48 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.25 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 135.60 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 123.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 131.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.02 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 101.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.98 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 95.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 139.77 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 126.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 125.15 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.12 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 148.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 120.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 131.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 128.05 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.28 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 131.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.20 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 123.13 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 115.19 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.42 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 135.81 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 128.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 111.32 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 126.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.45 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 98.59 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 131.51 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 119.52 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 126.86 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 113.16 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 112.88 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.36 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 104.71 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 91.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.70 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 121.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.95 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 136.29 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 124.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 132.18 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 118.30 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.75 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 109.25 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 95.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 151.58 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 126.46 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 136.73 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 118.72 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 147.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.41 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 132.14 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 146.31 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 118.84 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 128.66 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 110.97 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 141.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 127.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 126.85 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 143.21 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 113.49 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 123.91 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 105.62 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 137.86 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 124.56 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 123.44 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 144.55 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 114.39 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 124.94 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 106.65 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 140.92 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.64 KB
fmhaSm100fKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 124.24 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 106.01 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 94.80 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 111.07 KB
fmhaSm100fKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 100.29 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 102.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 87.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 105.13 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 91.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 102.08 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 90.97 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 105.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 93.93 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 112.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 98.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 115.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 101.53 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 147.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 151.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 163.96 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 138.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 144.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 153.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 152.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 165.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 140.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 144.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 134.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 148.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 159.04 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 127.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 142.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 140.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 149.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 161.08 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 143.53 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 142.92 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 134.04 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 123.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 145.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 130.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 162.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 145.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 128.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 135.39 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 145.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 136.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 126.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 123.97 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 149.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 134.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 164.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 150.43 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 126.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 138.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 124.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 132.65 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 136.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 147.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 123.44 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 129.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 138.25 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 139.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 152.00 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 125.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 132.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 119.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 133.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 142.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.46 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 125.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 136.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 147.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 130.42 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 123.72 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 117.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 107.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 101.92 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 133.60 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 117.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 147.46 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 132.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 111.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 95.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 121.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 126.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 123.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 110.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 108.43 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 137.37 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 151.74 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 136.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 113.19 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 98.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 124.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 128.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 140.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 121.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 132.78 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 145.93 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 119.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.56 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 112.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 126.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 136.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 130.36 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 141.15 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 124.01 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 115.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.58 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 101.60 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 95.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 126.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 109.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 140.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 125.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 103.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 89.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 114.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 100.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 118.69 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 104.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 130.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 145.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 130.72 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.67 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 92.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.91 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 104.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 125.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 128.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 140.49 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 121.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 132.13 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 132.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 145.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.61 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 112.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 125.87 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 135.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 119.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 130.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 141.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.58 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 124.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 115.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 109.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 101.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 95.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 124.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 138.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 122.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 88.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.61 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 118.11 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 116.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 104.11 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 102.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 129.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 112.04 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 143.81 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 127.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 106.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 92.06 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 103.95 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 143.23 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 128.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 146.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 131.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 124.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 112.25 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 127.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 115.21 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 116.64 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 105.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 119.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 108.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 115.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 104.01 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 118.28 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 107.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 155.50 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 154.78 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 168.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 143.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 147.57 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 157.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 156.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 170.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 145.60 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 148.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 142.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 152.40 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 163.51 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 133.18 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 146.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 144.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 153.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 165.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 134.76 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 147.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 154.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 141.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 134.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 124.38 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 149.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 134.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 167.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 150.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 133.36 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 117.08 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 140.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 125.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 157.03 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 141.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 137.40 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 128.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 153.44 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 138.24 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 170.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 154.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 131.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 119.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 143.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 127.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 140.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 140.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 152.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 133.63 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 142.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 144.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 156.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 130.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 136.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 127.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 137.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 147.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 131.81 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 129.65 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 141.59 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 151.78 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 120.09 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 134.99 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 135.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 118.83 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 109.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 137.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 119.00 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 152.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 136.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 116.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 100.37 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 128.53 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 110.79 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 138.52 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 128.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 121.64 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 112.94 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 141.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 122.16 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 156.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 140.17 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.28 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 103.23 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 131.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 113.67 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.35 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 133.33 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 145.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.73 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 125.99 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 136.82 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 137.48 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 150.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 129.88 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 130.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 140.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.56 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 123.61 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 135.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 145.74 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 114.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 128.33 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 127.61 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 112.89 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 103.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 129.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 112.26 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 144.70 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 129.50 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 108.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 94.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 120.83 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 104.04 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 130.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.41 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 115.82 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 134.54 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 150.66 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 134.44 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 111.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 97.75 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 124.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 133.11 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 132.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 145.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 125.53 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 136.83 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 137.27 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 150.81 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 124.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 129.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 120.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 130.31 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 141.07 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.91 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 123.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 123.62 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 134.72 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 146.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 114.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 128.12 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 127.00 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 117.05 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 112.93 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 128.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.90 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 143.85 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 128.02 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 108.45 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.20 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 117.22 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.80 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 129.84 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 120.98 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 115.86 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 107.34 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 133.55 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.30 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 148.77 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 133.10 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 111.68 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 97.50 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.29 KB
fmhaSm100fKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.82 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 110.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 98.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 112.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 101.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 109.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 101.81 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 112.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 104.65 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 155.09 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 156.12 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 156.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 154.24 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 160.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 151.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 151.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 146.24 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 155.06 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 147.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 163.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 154.93 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 154.13 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 145.59 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 152.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 141.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 130.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 127.34 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 135.17 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.68 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 137.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 132.70 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 124.87 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 119.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 131.77 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.77 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 140.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 136.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 127.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 128.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 133.18 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.03 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 117.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 124.18 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 124.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.23 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 113.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.00 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 105.15 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 127.20 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 126.11 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 116.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 121.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.10 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 159.61 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 150.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 162.47 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 153.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 139.51 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 129.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 142.37 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 132.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 125.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 116.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 128.19 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 119.91 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 162.64 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 161.54 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 160.71 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 159.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 160.77 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 153.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 160.19 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 152.28 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 157.61 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 150.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 157.82 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 145.97 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 137.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 132.64 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 139.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 134.97 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 140.88 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 126.89 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 136.91 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 123.44 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 141.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 132.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 138.32 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 124.76 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 124.97 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.81 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.69 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64MultiCtasKvGmemSepVarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 122.31 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 126.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 115.34 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 123.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 110.34 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 130.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.35 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 126.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512HVPerCta256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 231.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 222.70 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 228.55 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 219.04 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 227.26 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 241.71 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 219.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 224.60 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 240.36 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 216.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 202.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 192.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 225.25 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 214.06 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 190.93 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 184.95 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 198.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 194.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 225.10 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 216.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 186.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 183.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 193.23 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 183.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 199.09 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 189.58 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 188.81 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 205.06 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 181.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 194.90 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 210.04 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 187.11 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 177.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 167.43 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 202.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 189.52 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 164.48 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 155.68 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 180.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 167.90 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 205.57 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 190.41 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 166.37 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 158.20 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 178.15 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 168.89 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 183.59 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 174.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 173.72 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 189.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 166.31 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 179.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 195.65 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 171.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 163.20 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 150.70 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 186.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 172.43 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 149.56 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 138.91 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 169.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 156.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 193.27 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 179.00 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 155.67 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 144.87 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 236.36 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 226.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 232.96 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 223.05 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 231.78 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 246.46 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 224.12 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 229.13 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 244.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 221.08 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 206.82 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 196.84 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 230.72 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 221.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 195.57 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 188.40 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 203.07 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 198.62 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 235.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 220.74 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 191.29 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 186.59 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 197.63 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 188.12 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 203.62 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 193.72 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 193.33 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 209.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 185.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 199.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 214.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 191.25 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 182.23 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 171.68 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 206.62 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 192.90 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 170.85 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 159.25 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 184.16 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 172.02 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 212.98 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 194.38 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 172.51 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 161.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 182.66 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 173.03 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 187.99 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 178.21 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 178.24 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 194.05 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 170.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 183.71 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 200.39 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 175.53 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 167.18 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 154.83 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 190.55 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 176.13 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 155.94 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 142.49 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 173.30 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 160.45 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 201.80 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 182.48 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 161.52 KB
fmhaSm100fKernel_QkvFp16OFp16H512PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 148.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 102.50 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 90.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 109.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 98.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 101.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 90.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 109.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:34 97.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 113.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 101.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 120.36 KB
fmhaSm100fKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 108.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 123.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.74 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 117.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 111.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.86 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 133.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 111.42 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 122.33 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 114.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 107.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 115.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 104.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 102.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 123.81 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 110.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 113.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 101.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 94.78 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 118.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 106.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.74 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 102.17 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 87.50 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 89.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 122.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 110.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 126.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 114.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 125.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 91.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 105.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 93.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 113.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 99.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 130.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 107.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.87 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 111.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 102.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 111.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 101.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 98.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 121.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 116.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 102.61 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 110.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 98.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 102.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 90.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 115.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 103.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 116.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 99.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 84.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 97.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 85.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 119.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 107.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 106.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 94.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 123.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 111.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.72 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 89.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.24 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.36 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.35 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 129.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.67 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 110.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 109.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.39 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 109.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 96.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 104.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 114.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.78 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 88.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 114.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 102.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 101.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 83.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 95.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 83.37 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 117.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 104.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 121.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 110.03 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 119.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 106.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.01 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 87.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 99.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.81 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 103.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 109.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 99.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 131.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 107.82 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 118.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 112.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 111.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 102.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 101.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 121.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 106.86 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 116.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 105.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 102.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 110.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 97.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 101.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 89.54 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 115.28 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 103.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 102.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 84.45 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 84.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 118.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 106.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 93.82 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 122.69 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 110.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 120.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 107.61 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 103.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.28 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 88.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 126.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 105.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 115.53 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 134.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 117.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 104.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 114.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 100.94 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.00 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 100.28 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 103.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 92.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 114.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 99.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 87.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 119.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 107.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 123.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 102.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 112.08 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 98.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.10 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 100.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 110.44 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 97.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 122.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 109.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 97.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 99.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 87.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 110.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 96.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 95.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 83.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.40 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 104.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 121.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 100.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 109.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 96.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 130.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 112.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 98.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 108.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 95.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 120.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 96.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 97.21 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 85.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 108.33 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 93.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 93.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 81.35 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 114.86 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 103.45 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 123.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 101.78 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 111.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 98.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 114.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 100.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 110.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 97.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 122.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 96.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ16Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 98.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 87.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ32Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 109.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 96.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ8Kv256PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 93.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.03 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 116.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 106.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 125.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 114.21 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 114.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 103.83 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 122.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 111.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 112.81 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 102.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 121.08 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 110.03 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 113.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 102.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 122.10 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 110.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 130.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.44 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 123.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.27 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 139.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.01 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 127.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 120.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 121.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.45 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 110.58 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 129.57 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 115.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 125.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.98 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 120.61 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 108.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 110.87 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 130.07 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 117.83 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 124.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.47 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 107.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 92.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 105.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 93.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 127.76 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 115.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 114.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 102.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 137.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 124.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 129.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 117.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 111.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 96.64 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 109.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 96.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 127.85 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 118.90 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 115.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 103.85 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 136.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.42 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 124.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.49 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 118.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 117.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 107.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 126.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 122.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 110.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 106.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 117.74 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 105.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.37 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 94.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 127.59 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 115.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 120.65 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 108.22 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 100.94 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 89.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 125.13 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 112.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 110.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 98.67 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 134.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 122.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 125.89 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 108.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 93.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 105.48 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 126.32 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 116.73 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 101.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 134.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 110.25 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 121.87 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 116.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 105.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 117.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.97 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 114.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.72 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.26 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 109.41 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 120.08 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 109.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 116.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.02 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 92.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 126.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 114.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 118.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 106.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 103.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 88.56 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.51 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 86.82 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 123.61 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 111.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 108.33 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 96.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 132.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 120.62 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 123.58 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 92.39 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 90.70 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 128.30 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.75 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.03 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 115.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 136.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 112.46 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 123.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 118.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 107.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 119.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 106.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 108.05 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 127.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 122.29 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 110.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 106.84 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 105.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 105.51 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 93.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 127.76 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 115.72 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 119.95 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 107.14 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.55 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 90.06 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 99.99 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 88.52 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 125.24 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 113.15 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 109.81 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ16Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 97.96 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 134.34 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 122.31 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 125.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ32Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 112.50 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 108.69 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ64Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 93.93 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 104.66 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ8Kv256StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 92.29 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 134.04 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 111.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 123.11 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 106.43 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 133.16 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 123.12 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 122.20 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 131.23 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 106.85 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 118.77 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 102.09 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 130.42 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.29 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 119.90 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 129.69 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 104.68 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 116.60 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 99.92 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 128.88 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 118.76 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 118.38 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 131.79 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 106.80 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 118.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv256StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 101.91 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 132.71 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv256StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.35 KB
fmhaSm100fKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 119.95 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 106.77 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 90.05 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 113.19 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 97.20 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 107.05 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 89.79 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 114.39 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 96.90 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 117.44 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 100.97 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 123.89 KB
fmhaSm100fKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 108.02 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 128.33 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 111.70 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 134.88 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 118.73 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.66 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 106.33 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 129.08 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 113.36 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 119.42 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 102.82 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 125.84 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 109.85 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 128.97 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 111.31 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 136.27 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 118.47 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 122.34 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 105.65 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 129.77 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 112.68 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 119.22 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 102.70 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 126.66 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 109.73 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 139.87 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 123.45 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 146.31 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 129.34 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 134.20 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 118.06 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 140.63 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 124.09 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 130.84 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 114.43 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 137.27 KB
fmhaSm100fKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 120.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 91.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 79.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 87.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 85.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 74.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 100.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 89.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 96.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 94.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 84.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 85.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 78.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 84.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 80.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 74.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 94.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 94.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 89.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 84.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 78.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 70.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 74.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 82.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 72.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 73.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 64.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 67.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 59.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 84.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 75.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 80.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 69.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 88.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 79.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 79.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 71.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 73.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 66.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 77.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 84.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 82.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 72.67 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 98.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 87.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 94.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 82.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 83.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 75.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 78.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 72.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 92.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 85.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 92.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 87.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 81.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 76.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 67.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 60.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 79.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 70.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 71.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 61.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 64.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 56.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 81.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 73.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 78.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 67.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 86.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 77.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 77.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 68.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 63.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 87.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 76.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 83.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 81.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 71.42 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 97.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 85.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 93.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 91.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 81.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 82.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 74.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 80.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 76.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 70.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 91.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 84.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 90.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 86.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 80.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 74.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 66.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 70.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 59.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 78.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 69.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 69.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 60.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 63.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 55.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 80.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 76.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 66.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 85.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 76.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 76.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 67.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 70.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 62.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 89.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 77.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 85.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 73.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 98.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 87.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 94.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 83.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 83.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 76.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 82.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 78.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 72.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 93.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 86.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 92.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 88.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 82.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 75.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 67.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 71.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 60.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 79.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 70.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 70.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 61.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 64.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 56.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 80.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 73.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 77.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 67.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 85.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 77.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 77.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 68.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 71.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 63.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 94.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 78.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 85.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 74.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 103.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 88.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 95.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 84.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 77.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 83.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 73.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 97.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 87.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 93.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 83.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 76.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 71.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 69.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 64.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 77.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 66.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 61.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 82.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 80.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 75.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 74.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 83.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 81.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 72.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 92.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 76.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 83.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 72.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 101.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 93.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 81.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 86.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 75.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 81.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 71.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 95.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 91.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 81.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 74.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 69.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 66.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 61.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 74.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 63.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 59.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 80.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 78.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 73.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 81.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 79.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 70.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 91.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 75.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 82.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 100.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 84.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 92.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 80.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 85.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 73.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 79.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 93.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 84.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 89.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 79.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 65.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 60.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 73.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 68.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 62.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 58.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 79.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 77.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 72.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 80.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 78.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 69.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 67.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 76.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 84.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 101.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 86.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 93.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 82.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 86.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 75.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 81.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 95.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 91.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 81.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 73.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 69.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 66.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 61.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 74.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 69.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 63.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 59.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 79.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 78.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 73.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 81.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 79.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 70.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 78.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 73.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 88.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 83.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 76.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 86.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 82.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 68.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 64.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 65.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 61.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 75.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 74.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 72.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 71.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 76.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.53 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 81.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 74.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 70.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 84.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 80.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 65.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 61.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 63.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 59.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 72.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 69.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 74.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 70.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 80.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 73.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 83.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 79.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 64.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 60.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 61.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 58.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 71.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 70.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 68.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 76.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 72.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 86.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 82.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 75.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 71.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 85.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 81.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 65.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 61.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 62.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 59.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 72.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 69.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 98.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 83.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 81.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 90.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 78.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 76.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 107.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 92.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 91.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 100.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 87.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 92.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 81.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 79.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 88.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 85.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 77.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 76.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 91.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 89.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 98.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 94.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 87.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 85.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 85.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 76.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 77.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 73.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 66.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 62.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 86.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 78.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 69.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 72.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 69.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 62.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 59.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 90.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 82.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 83.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 80.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 73.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 69.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 92.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 84.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 76.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 78.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 75.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 68.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 66.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 96.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 80.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 79.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 89.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 88.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 75.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 74.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 105.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 90.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 89.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 97.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 85.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 84.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 90.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 79.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 77.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 86.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 83.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 75.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 73.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 89.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 87.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 96.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 92.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 85.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 83.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 83.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 74.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 71.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 63.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 60.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 84.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 74.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 76.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 66.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 69.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 67.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 59.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 57.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 80.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 81.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 77.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 71.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 67.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 90.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 81.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 73.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 76.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 73.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 66.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 64.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 94.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 79.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 77.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 87.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 86.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.67 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 73.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 89.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 87.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 97.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 96.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 84.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 82.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 89.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 78.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 76.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 84.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 81.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 73.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 72.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 98.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 88.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 86.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 94.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 91.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 83.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 82.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 81.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 73.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 73.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 69.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 59.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 83.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 73.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 74.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 65.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 68.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 65.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 58.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 56.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 87.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 79.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 80.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 76.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 69.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 66.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 89.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 80.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 80.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 72.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 75.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 72.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 65.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 63.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 96.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 79.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 89.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 76.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 75.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 91.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 89.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 97.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 86.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 85.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 90.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 80.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 78.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 86.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 83.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 75.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 74.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 100.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 90.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 88.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 93.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 85.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 84.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 83.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 74.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 74.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 69.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 64.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 60.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 84.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 74.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 76.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 67.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 68.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 66.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 60.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 57.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 80.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 81.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 76.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 71.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 67.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 90.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 81.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 74.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 74.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 72.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 67.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 64.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 115.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 92.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 100.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 86.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 104.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 96.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 89.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 97.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 84.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 102.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 94.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 111.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 88.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 96.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 83.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 101.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 93.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 114.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 90.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 98.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 85.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 102.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 94.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 98.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 108.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 95.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 119.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 127.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 126.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 115.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 97.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 94.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 125.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 117.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 124.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 120.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 114.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 93.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 81.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 104.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 92.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 94.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 85.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 77.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 111.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 108.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 92.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 118.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 103.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 108.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 95.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 100.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 88.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 93.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 102.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 101.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 89.67 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 113.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 109.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 91.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 98.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 95.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 88.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 119.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 111.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 118.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 115.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 108.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 87.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 76.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 86.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 78.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 81.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 71.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 106.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 94.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 101.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 88.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 102.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 90.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 83.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 90.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 98.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 98.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 87.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 111.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 119.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 118.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 107.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 89.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 95.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 92.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 86.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 117.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 109.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 116.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 112.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 106.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 84.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 73.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 84.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 85.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 76.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 79.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 69.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 103.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 92.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 99.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 85.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 109.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 96.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 100.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 88.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 91.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 91.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 99.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 88.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 120.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 119.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 108.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 90.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 93.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 87.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 110.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 113.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 107.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 84.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 74.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 94.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 84.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 86.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 76.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 69.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 103.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 93.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 99.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 86.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 109.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 97.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 100.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 89.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 92.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 82.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 97.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 94.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 118.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 114.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 96.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 93.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 116.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 113.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 90.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 82.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 83.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 79.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 104.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 103.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 98.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 100.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 92.31 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 89.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 112.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 109.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 90.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 87.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 111.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 83.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 77.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 77.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 74.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 98.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 98.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 92.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 94.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 89.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.53 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 110.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 106.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 88.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 85.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 105.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 81.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 74.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 75.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 71.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 95.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 95.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 89.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 92.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 91.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 87.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 111.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 108.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 89.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 86.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 109.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 106.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 81.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 75.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 75.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 96.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 95.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 90.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 92.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 100.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 112.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 112.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 99.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 97.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 123.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 121.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 132.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 132.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 119.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 117.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 101.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 99.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 108.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 106.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 97.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 96.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 132.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 121.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 119.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 129.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 126.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 118.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 116.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 93.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 84.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 81.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 108.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 96.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 100.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 89.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 92.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 88.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 80.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 78.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 106.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 111.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 107.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 96.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 92.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 122.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 108.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 113.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 100.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 106.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 103.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 91.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 89.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 97.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 95.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 106.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 106.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 93.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 91.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 117.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 115.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 126.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 126.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 113.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 111.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 95.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 93.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 102.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 100.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 92.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 90.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 127.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 113.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 123.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 120.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 112.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 110.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 90.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 86.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 79.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 75.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 101.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 91.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 93.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 83.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 85.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 82.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 74.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 72.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 113.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 101.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 105.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 101.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 91.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 87.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 116.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 107.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 95.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 96.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 86.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 84.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 95.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 92.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 103.81 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 91.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 89.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 115.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 113.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 124.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 124.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 111.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 109.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 93.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 90.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 100.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 89.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 88.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 124.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 113.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 111.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 120.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 118.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 109.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 108.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 88.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 84.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 76.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 99.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 88.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 90.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 81.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 82.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 79.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 72.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 70.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 110.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 99.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 102.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 98.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 89.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 85.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 113.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 101.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 104.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 93.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 97.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 94.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 84.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 82.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 95.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 94.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 104.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 105.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 92.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 90.53 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 116.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 114.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 124.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 125.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 112.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 110.67 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 94.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 92.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 101.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 99.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 90.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 89.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 125.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 114.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 112.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 122.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 119.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 111.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 109.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 88.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 85.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 77.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 74.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 99.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 89.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 91.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 82.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 81.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 79.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 73.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 111.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 100.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 103.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 99.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 90.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 86.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 114.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 101.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 105.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 94.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 95.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 93.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 85.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 83.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 184.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 179.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 181.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 178.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 150.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 167.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 145.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 163.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 171.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 167.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 168.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 165.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 136.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 155.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 131.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 151.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 165.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 161.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 163.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 159.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 132.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 149.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 126.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 146.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 187.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 183.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 185.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 181.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 154.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 143.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 148.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 139.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 174.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 170.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 172.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 168.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 139.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 129.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 135.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 125.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 169.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 164.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 166.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 162.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 135.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 125.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 130.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H512PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 121.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 81.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 68.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 75.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 75.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 66.42 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 86.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 74.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 80.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 80.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 77.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 67.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 74.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 70.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 65.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 82.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 73.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 79.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 76.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 71.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 72.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 63.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 64.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 56.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 74.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 63.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 64.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 55.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 61.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 53.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 76.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 67.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 68.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 60.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 78.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 67.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 59.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 65.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 57.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 79.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 66.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 73.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 72.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 64.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 84.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 71.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 78.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 77.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 69.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 74.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 65.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 71.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 68.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 63.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 79.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 71.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 77.15 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 73.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 68.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 70.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 61.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 61.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 53.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 72.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 60.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 62.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 53.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 58.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 50.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 73.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 65.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 66.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 58.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 76.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 65.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 65.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 57.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 62.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 55.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 78.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 72.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 71.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 62.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 70.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 77.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 76.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 68.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 73.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 64.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 70.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 67.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 62.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 78.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 69.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 76.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 72.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 67.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 69.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 59.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 60.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 52.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 71.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 59.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 60.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 52.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 57.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 49.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 72.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 64.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 64.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 57.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 75.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 64.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 64.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 56.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 61.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 53.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 79.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 66.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 73.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 73.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 64.92 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 85.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 72.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 79.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 78.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 70.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 75.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 66.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 72.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 69.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 64.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 80.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 71.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 77.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 74.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 69.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 69.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 61.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 61.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 53.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 71.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 60.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 61.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 53.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 58.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 50.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 72.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 65.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 66.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 58.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 75.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 65.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 65.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 57.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 62.90 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 55.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 84.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 67.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 74.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 65.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 89.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 73.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 79.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 71.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 79.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 66.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 72.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 78.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 70.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 70.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 63.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 62.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 55.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 69.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 60.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 60.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 53.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 74.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 68.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 66.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 60.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 73.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 65.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 64.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 58.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 82.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 72.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 63.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 87.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 77.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 77.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 64.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 70.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 62.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 82.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 70.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 76.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 60.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 60.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 52.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 67.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 58.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 58.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 50.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 72.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 66.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 64.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 58.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 71.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 63.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 62.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 56.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 81.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 64.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 62.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 86.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 76.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 67.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 76.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 63.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 69.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 61.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 81.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 75.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 66.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 67.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 59.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 58.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 51.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 66.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 56.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 56.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 49.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 71.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 64.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 63.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 57.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 70.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 62.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 60.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 54.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 82.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 66.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 71.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 78.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 69.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 78.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 65.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 63.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 83.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 76.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 68.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 67.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 60.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 59.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 52.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 67.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 57.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 57.65 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 50.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 71.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 65.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 63.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 58.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 71.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 63.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 61.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvCustomPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 55.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 65.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 72.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 70.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 66.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 69.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 61.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 59.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 53.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 65.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 60.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 63.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 58.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 63.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 64.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 62.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 69.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 59.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 52.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 57.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 50.83 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 63.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 58.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 61.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 56.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 63.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 61.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 69.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 62.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 61.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.37 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 66.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 57.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 51.61 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 55.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 49.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 62.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 56.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 59.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDenseP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 54.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 65.56 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 63.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 64.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 63.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 68.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 59.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 52.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 57.02 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 50.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 63.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 57.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 61.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvDensePagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 55.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 88.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 72.20 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 70.19 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 79.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 80.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 69.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 68.28 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 94.11 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 77.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 75.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 84.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 85.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 75.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 73.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 71.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 69.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 78.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 76.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 69.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 67.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 89.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 76.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 74.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 83.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 81.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 74.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 72.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 79.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 70.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 70.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 66.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 59.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 55.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 78.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 67.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 71.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 60.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 66.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 63.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 56.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 53.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 83.09 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 74.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 74.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 70.39 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 63.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 60.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 82.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 71.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 75.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 64.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 70.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 67.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 60.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 58.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 86.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 69.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 67.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 77.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 77.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 67.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 66.03 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 91.98 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 75.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 73.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 82.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 82.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 72.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 81.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 69.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 67.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 76.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 73.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 66.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 65.33 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 87.18 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 74.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 72.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 81.52 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 79.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 72.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 70.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 76.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 68.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 67.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 64.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 57.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 53.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 76.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 69.05 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 57.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 63.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 61.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 53.96 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 51.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 80.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 72.41 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 71.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 68.14 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 61.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 58.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 80.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 69.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 72.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 62.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 67.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 64.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 58.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 55.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 85.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 68.70 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 66.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 76.34 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 76.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 66.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 64.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 90.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 74.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 72.07 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 81.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 81.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 71.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 70.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 80.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 68.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 65.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 74.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 72.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 64.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 85.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 73.51 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 71.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 80.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 77.76 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 70.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 69.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 75.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 66.97 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 66.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 62.79 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 55.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 52.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 75.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 63.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 67.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 56.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 62.64 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 59.75 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 52.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 50.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 79.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 71.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 70.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 66.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 60.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 57.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 79.30 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 68.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 71.66 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 60.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 66.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 63.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 57.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 54.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 86.91 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 70.58 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.69 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 78.22 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 78.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 68.17 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 66.78 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 92.73 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 76.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 74.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 83.72 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 83.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:25 73.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.29 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 82.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 70.00 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 67.86 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:31 76.89 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 74.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 67.47 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 66.08 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 88.06 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:28 75.63 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 82.40 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 79.88 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 72.85 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 77.16 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 68.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 65.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 62.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 57.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 54.10 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 76.44 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 65.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 66.68 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 58.24 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 62.27 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 60.25 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 54.21 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqKOnlyTransQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 51.94 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 80.71 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:29 70.04 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 66.77 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:33 61.99 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 58.60 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:27 80.43 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ32Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:26 69.50 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 70.54 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ64Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 62.49 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentGroupedSwapsAbForGen.cubin 22-Jul-2026 16:30 66.38 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 64.36 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticGroupedSwapsAbForGen.cubin 22-Jul-2026 16:32 58.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 56.32 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 102.82 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 77.12 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 85.26 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 73.84 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 96.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 87.93 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.57 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 74.87 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 83.13 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.59 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 94.48 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 85.80 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 99.45 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 73.62 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 82.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 70.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 93.23 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 84.55 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 101.95 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 75.74 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 84.01 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqSeparateTransKVQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 72.46 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 94.35 KB
fmhaSm103aKernel_QBfloat16KvE4m3OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqSeparateTransKVQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 86.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 93.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 86.07 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 91.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 83.82 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 89.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 81.70 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 89.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 81.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 93.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 87.06 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 91.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 84.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 89.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 82.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 89.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 82.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 100.47 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 93.00 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 98.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 90.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 96.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 88.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 97.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 89.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 105.39 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 92.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 97.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 84.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 93.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 80.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 93.64 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 80.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 105.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 99.89 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 98.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 92.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 94.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 88.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 94.14 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 88.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 112.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 98.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 104.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 91.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 100.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 87.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 101.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 87.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 88.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 82.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 84.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 76.61 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 82.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 74.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 83.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 74.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 87.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 83.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 84.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 77.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 82.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 75.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 83.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 75.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 95.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 88.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 91.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 83.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 89.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 81.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 90.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 82.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 96.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 88.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 94.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 86.53 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 92.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 84.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 92.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 84.61 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 95.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 73.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 80.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 69.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 97.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 88.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 72.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 78.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 90.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 77.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 69.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 67.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 58.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 74.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 65.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 64.65 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 56.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 79.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 93.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 71.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 78.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 67.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 95.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 85.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 76.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 66.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 74.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 67.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 64.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 56.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 72.45 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 63.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 62.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 54.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 76.95 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 70.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 91.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 70.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 77.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 65.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 93.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 84.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 68.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 74.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 64.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 86.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 72.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 65.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 62.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 55.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 69.45 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 62.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 60.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 53.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 74.20 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 68.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 92.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 78.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 66.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 94.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 85.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 70.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 75.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 66.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 87.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 72.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 66.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 62.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 55.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 69.95 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 62.65 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 60.65 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 53.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 74.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 68.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 73.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 68.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 72.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 68.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 65.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 58.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 96.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 89.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 63.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 56.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 66.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 69.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 66.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 63.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 56.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 94.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 86.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 61.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 69.71 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 68.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 64.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 61.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 55.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 92.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 84.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 59.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 52.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 71.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 66.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 65.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 62.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 55.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 92.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 85.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 60.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 53.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 76.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 71.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 74.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 70.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 71.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 60.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 103.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 95.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 67.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 58.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 74.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 72.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 68.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 58.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 101.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 93.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 65.50 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 56.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 72.71 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 67.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:28 66.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 56.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 99.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 91.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 63.00 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 54.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 74.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 69.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 72.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 68.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 65.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 57.61 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 100.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 92.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 63.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 55.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 106.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 77.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 85.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 100.89 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 92.47 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 104.62 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 74.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 82.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 69.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 89.97 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 102.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 73.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 81.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 68.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 96.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 88.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 104.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 74.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 82.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 97.64 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H128PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 89.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 108.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 94.50 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 100.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 87.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 96.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 82.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 96.64 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 83.12 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 98.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 94.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 96.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 93.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 108.53 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 102.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 90.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 82.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 87.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 79.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 91.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 87.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 89.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 86.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 101.03 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 95.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 81.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 75.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 79.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 73.23 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 86.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 83.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 85.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 82.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 97.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 90.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 77.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 75.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 87.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 84.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 85.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 82.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 97.14 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 90.39 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 78.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 71.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 75.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 101.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 97.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 99.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 96.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 115.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 101.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 94.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 83.86 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 90.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 81.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 94.71 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 90.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 92.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 89.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 107.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 94.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 87.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 76.36 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:26 84.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 73.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 90.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 86.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 88.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 84.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 103.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 89.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 83.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 80.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 90.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 87.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 89.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 85.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 104.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 89.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 81.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 72.36 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 78.50 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.96 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 87.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 82.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 84.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 76.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 82.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 74.36 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 82.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 74.86 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 90.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 72.21 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 78.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 70.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.84 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 85.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 71.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 77.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 69.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 87.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 74.97 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 67.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 66.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 59.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 71.07 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 64.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 64.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 57.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 77.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP16VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 69.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 86.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 67.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 74.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 65.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 88.34 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 80.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 66.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 73.02 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 64.98 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 82.92 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 70.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 63.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 62.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 55.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 67.57 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 60.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 60.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 53.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 72.84 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP32VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 65.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 84.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 66.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 72.72 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 64.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 86.59 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 79.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 65.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 71.40 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 63.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64MultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 81.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 68.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:27 61.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 60.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 53.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 65.20 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 58.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 58.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 51.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:25 70.22 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomP64VarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 63.67 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 85.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 67.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 73.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 65.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvCgaVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 87.84 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 80.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 66.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 72.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 64.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedMultiCtasKvVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 82.42 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 68.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 62.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 61.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 65.95 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 59.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:29 58.90 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 52.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 70.97 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvCustomPagedVarSeqSkipsSoftmaxQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 64.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 69.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.10 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 64.80 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 59.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 87.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 82.93 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 62.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:25 57.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 67.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.30 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 66.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 64.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 61.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 55.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 84.54 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 76.87 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 59.52 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 53.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 65.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 63.68 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 64.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 62.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 59.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 53.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 82.04 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 74.75 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:25 57.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDenseP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 51.60 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 65.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 66.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 64.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 60.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 54.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 82.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 75.37 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 58.27 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvDensePagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 52.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 74.58 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 72.55 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 73.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 71.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 70.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 59.99 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 94.78 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 87.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:31 67.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 57.83 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 70.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 68.05 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 69.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 67.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 66.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 56.24 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 91.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 83.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 63.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 54.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.46 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 66.43 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:29 67.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 65.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:32 63.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 54.24 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 88.91 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 81.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:27 60.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 52.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 70.08 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvCgaVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 68.18 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 69.26 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 67.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:33 63.66 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 55.49 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 90.28 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 82.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128PersistentSwapsAbForGen.cubin 22-Jul-2026 16:30 61.25 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:32 53.33 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 104.13 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 75.88 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 84.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 72.85 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 98.29 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP16VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 89.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:27 71.38 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:28 79.65 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.35 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 95.16 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP32VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 85.61 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:26 98.01 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 69.76 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:26 78.15 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64MultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:30 66.73 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 92.41 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomP64VarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 83.74 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 99.63 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ16Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:31 71.51 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ32Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 79.77 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedMultiCtasKvVarSeqQ8Kv128StaticSwapsAbForGen.cubin 22-Jul-2026 16:33 68.48 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128PersistentGroupedKeepsAbForGen.cubin 22-Jul-2026 16:33 93.79 KB
fmhaSm103aKernel_QE4m3KvE2m1OE4m3H64PagedKvSlidingWindowCustomPagedVarSeqQ128Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 84.74 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 105.54 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 94.05 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 109.53 KB
fmhaSm103aKernel_QkBfloat16VE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 98.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 105.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 94.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 109.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 98.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 106.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 95.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 109.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 99.11 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 116.03 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 104.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 119.61 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 108.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 127.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 116.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 131.14 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 120.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 122.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 110.99 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 125.99 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 114.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 118.25 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 106.98 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 121.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 110.81 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 118.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 107.83 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 122.08 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 111.53 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 127.96 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 117.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 131.72 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 121.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 123.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 111.77 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 126.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 115.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 118.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 108.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 122.41 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 111.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 119.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 108.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 122.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 112.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 138.90 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 128.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 142.44 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 131.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 133.76 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 122.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 137.43 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 126.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 129.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 118.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 133.25 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 122.20 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 131.13 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 119.68 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 134.67 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 123.40 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 129.62 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 124.32 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:31 121.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:32 121.72 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 105.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 94.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 109.15 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 98.18 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 106.15 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 95.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 110.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 99.11 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 101.27 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 86.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 103.98 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 89.96 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 100.93 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 89.82 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 104.03 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 92.77 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 111.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 97.40 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 114.60 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 100.67 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 141.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 122.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 144.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 125.09 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 122.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 106.40 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 125.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 109.19 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 114.65 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 100.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 117.65 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 103.24 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 114.01 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 100.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 116.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 102.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 142.21 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 127.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 145.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 130.00 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 123.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 111.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 126.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 114.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 115.50 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 104.07 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 118.47 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 107.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 114.14 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 102.86 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 117.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 105.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 153.03 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 133.59 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 156.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 136.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 134.39 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 117.69 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 137.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 120.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 126.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 111.74 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 129.41 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 114.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 125.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 111.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 128.67 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 114.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 100.98 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 86.96 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 104.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 89.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 101.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 89.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 104.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 92.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 110.00 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 98.27 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 112.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 101.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 109.69 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 101.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 112.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 104.37 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 160.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 151.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 163.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 153.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 137.72 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 124.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 140.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:30 127.64 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 123.82 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 113.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 127.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 116.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 159.47 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 153.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 162.44 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 156.43 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 139.37 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 128.68 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 142.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 131.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 125.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 116.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 127.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 119.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 100.50 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 88.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 106.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 94.14 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 100.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 89.55 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 106.66 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 95.06 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 111.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 99.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 116.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 105.13 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 116.40 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 104.53 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 122.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 110.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 114.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 102.26 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 119.35 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 107.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 112.50 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 100.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 117.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 106.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 113.28 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 101.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 118.84 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 106.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 116.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 105.38 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 122.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 111.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 114.56 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 103.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 119.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:25 108.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 113.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 101.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 118.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 107.27 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 114.05 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 102.24 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 119.38 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 107.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 127.94 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 115.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 134.01 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 121.79 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 125.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 113.36 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 131.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 119.01 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 123.92 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 111.95 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 129.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 117.61 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 125.72 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 113.65 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 131.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 119.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:30 120.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:29 118.20 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 116.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:28 118.24 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 104.12 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 88.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 109.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 94.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 104.52 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 89.37 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 110.28 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 95.02 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 114.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 99.45 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 120.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 104.91 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 125.92 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 110.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 131.54 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 115.63 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 120.13 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 104.93 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 125.73 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 110.38 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 116.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 101.30 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 122.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 106.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 126.11 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 110.89 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 132.09 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 116.46 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 119.60 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 105.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 125.58 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 110.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 116.48 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 102.15 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 122.47 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 107.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 137.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 121.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 142.85 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 127.08 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 131.67 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 116.04 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 137.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 121.70 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 128.31 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 112.41 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 133.80 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 118.20 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 107.56 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 96.38 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 112.75 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 101.42 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 108.36 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 97.10 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 113.71 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 102.34 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 112.49 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 99.61 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 116.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 103.23 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 113.29 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 101.88 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 117.16 KB
fmhaSm103aKernel_QkvBfloat16OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 105.70 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 101.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 90.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 105.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 94.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 102.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:29 112.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:27 101.16 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:28 141.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:31 129.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:30 114.28 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:27 103.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 91.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 106.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:28 115.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:27 104.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:33 144.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:31 133.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:29 118.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:27 106.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 95.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 112.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 101.32 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 116.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 105.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 118.33 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 106.66 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 122.01 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 110.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 115.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 103.85 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 119.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 107.71 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 113.83 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 102.32 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 117.51 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 106.17 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 115.00 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 103.94 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:27 118.68 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 107.79 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 118.71 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 107.88 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 122.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 111.91 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 115.88 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 105.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 119.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 108.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 114.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 103.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:32 118.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 107.57 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 115.63 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 104.85 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:30 119.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:30 108.90 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 130.28 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 119.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:26 133.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 122.04 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 127.33 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:28 116.15 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 130.48 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:32 119.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 125.79 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:26 114.74 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 128.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 117.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:31 128.11 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 116.16 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 131.14 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:27 119.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:25 102.08 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:33 90.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:33 105.76 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 94.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:29 102.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:28 112.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:29 101.16 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:32 141.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:33 129.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:29 114.41 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:27 103.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:31 91.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:28 106.51 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:27 116.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:25 104.92 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:26 145.07 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:32 133.90 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:33 118.18 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:33 106.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:29 95.57 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 95.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 84.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 98.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 87.11 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:25 97.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 85.42 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 99.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 88.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 108.00 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 95.63 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 110.67 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 98.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 118.61 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 106.67 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 120.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 111.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 100.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 102.36 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 97.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.89 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:27 99.29 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 110.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.28 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:32 100.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:26 121.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:28 107.43 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:33 123.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.09 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.01 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.79 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 103.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.88 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:29 98.07 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.58 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 100.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.31 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:26 98.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.15 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.18 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 132.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:31 118.47 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 135.31 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 120.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:32 124.60 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:25 111.87 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 114.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:27 121.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.02 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 111.44 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.64 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.82 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.12 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:28 97.52 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 84.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:30 98.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:33 87.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 97.48 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 85.55 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:29 100.05 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H256SeparateQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 88.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:31 105.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 93.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.75 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 95.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 94.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.97 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 97.06 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 135.87 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 123.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 138.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 125.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.79 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 111.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.18 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 113.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.91 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 137.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 123.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 140.17 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 126.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.31 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.65 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.14 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.44 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.93 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.66 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 95.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 83.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 99.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 87.57 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 96.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 111.68 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 89.00 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 139.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 117.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 113.68 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 90.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 84.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 100.46 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 112.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 92.58 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 141.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 121.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 94.64 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 88.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.20 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.46 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.35 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.48 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.64 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.46 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.61 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.40 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.81 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.44 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.84 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.41 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.72 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.70 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.69 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.72 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.63 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.87 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.63 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.95 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.78 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.82 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.24 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.29 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.20 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.33 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.23 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.45 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.27 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.76 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 115.59 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.30 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.47 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.92 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.06 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.26 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.41 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.91 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.67 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 111.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 89.00 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 139.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 117.91 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 113.80 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 90.94 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 112.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 92.70 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 141.39 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 121.51 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.58 KB
fmhaSm103aKernel_QkvE4m3OBfloat16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 94.77 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 99.34 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 87.36 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 103.28 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 91.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 99.86 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 88.04 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 103.98 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk128HV64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 92.02 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.66 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.38 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.21 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.10 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.22 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.62 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.96 KB
fmhaSm103aKernel_QkvE4m3OBfloat16HQk192HV128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.67 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 103.25 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 91.83 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.56 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 103.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 102.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 143.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 131.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 116.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 104.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 92.73 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 118.17 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 106.54 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 147.14 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 135.64 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 120.25 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 108.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.77 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.98 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.63 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.57 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.09 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.58 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.06 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.13 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.29 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.00 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.18 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.22 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.23 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.68 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.96 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.98 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.90 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 131.87 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 120.27 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 134.90 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 123.47 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 117.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.07 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 120.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 116.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.53 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 119.24 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.44 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 117.47 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 120.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 104.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 102.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 143.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 131.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 116.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 104.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 92.85 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 118.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 106.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 147.14 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 135.64 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 120.25 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 108.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.77 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 96.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 85.54 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 99.77 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 87.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 98.29 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 86.24 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 100.99 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 88.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.06 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 96.34 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.73 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 99.17 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.55 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.49 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.54 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 102.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 97.97 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 99.98 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.41 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.22 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 100.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.35 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.12 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.02 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.84 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.73 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 104.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.89 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.43 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 99.22 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.09 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 102.01 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 133.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 119.16 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 136.44 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 121.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.08 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.21 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.13 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.07 KB
fmhaSm103aKernel_QkvE4m3OE2m1H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 113.02 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.72 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 94.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.56 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 96.55 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.34 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 95.07 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 97.63 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 136.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 123.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 139.31 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 125.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.74 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.24 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 114.65 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.87 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 138.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 124.36 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 141.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 126.92 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 113.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.09 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.66 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.21 KB
fmhaSm103aKernel_QkvE4m3OE2m1H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 111.23 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 100.16 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 84.03 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 103.75 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 88.02 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 100.78 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 101.81 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 90.43 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 130.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 119.21 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 103.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 92.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 84.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 104.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 105.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 94.27 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 134.36 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 122.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 107.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 96.33 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 88.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.70 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.32 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.61 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.71 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.83 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.15 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.15 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.99 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.05 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.59 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.09 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.88 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.98 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.45 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.88 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.14 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.13 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.60 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.33 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.77 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.80 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.76 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.12 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.30 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 131.46 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 115.63 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.02 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.50 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.48 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.10 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.93 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.43 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.04 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.38 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.52 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.71 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 101.94 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 90.55 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 130.62 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 119.34 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 103.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 92.37 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 105.82 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 94.39 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 134.36 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 122.95 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 107.91 KB
fmhaSm103aKernel_QkvE4m3OE2m1H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 96.33 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.82 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.21 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.06 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.49 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 125.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 114.23 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 154.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 142.92 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 127.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 116.17 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.23 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.38 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 129.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 117.88 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 157.90 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 147.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 131.00 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 119.94 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.83 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.81 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.30 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 116.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.95 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.66 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.15 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.56 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.70 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.42 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.51 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 117.42 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.69 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.79 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 114.61 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.24 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.08 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.53 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.41 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 114.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 136.43 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 124.66 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 139.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 127.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 133.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 121.96 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 136.51 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 125.03 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 131.94 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 120.30 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 134.97 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 123.50 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 134.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 121.84 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 137.29 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 125.04 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 125.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 114.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 154.28 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 142.92 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 127.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 116.17 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.50 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 129.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 118.00 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 158.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 147.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 131.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 119.94 KB
fmhaSm103aKernel_QkvE4m3OE4m3H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 101.33 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 89.90 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 104.18 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 92.04 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.70 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 90.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 105.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 93.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 100.57 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.14 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 103.41 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.95 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 111.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 113.74 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.95 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 105.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.29 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 102.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 104.23 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.82 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 103.03 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 105.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.02 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.42 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 106.07 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.61 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 103.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.93 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 105.66 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.79 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 103.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.49 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 106.24 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 137.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 123.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 140.77 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 126.03 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.07 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 116.68 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.70 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 119.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 113.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 116.37 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.11 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 114.76 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 117.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.60 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.91 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 100.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 99.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 141.93 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 128.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 144.30 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 130.11 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.85 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 116.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 118.82 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 114.10 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 142.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 128.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 145.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 131.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.72 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 117.50 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 133.43 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 120.10 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.84 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.38 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 95.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 83.88 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 99.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 87.86 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 96.18 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 105.48 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 93.62 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 134.19 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 122.40 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 107.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 95.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 84.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 100.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 109.37 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 97.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 138.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 126.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 111.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 99.51 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 88.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.35 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.57 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.36 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.79 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.64 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.25 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.96 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.73 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.71 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.44 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.72 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.74 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.70 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.72 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.72 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.87 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.77 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.76 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.98 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.78 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.09 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.22 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.24 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.20 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.60 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.10 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.59 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.14 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.76 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 115.47 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.18 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.34 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.67 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.80 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.64 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.81 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.13 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.27 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.54 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.22 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.91 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 105.61 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 93.74 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 134.31 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 122.52 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 107.45 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 95.55 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 109.49 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 97.57 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 138.05 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 126.12 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 111.46 KB
fmhaSm103aKernel_QkvE4m3OE4m3H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 99.51 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.20 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 90.88 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 105.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 94.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.88 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 112.45 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 101.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 141.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 129.97 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 103.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 91.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 116.10 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 104.92 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 145.08 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 133.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 118.18 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 106.98 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.57 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.70 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.38 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.75 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.43 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.09 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.21 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.70 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.89 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.38 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.31 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.93 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.17 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.09 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.25 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.84 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.26 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.44 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.95 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.88 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.11 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 119.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 133.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 122.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.71 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 116.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.87 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 119.73 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 115.00 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.33 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 118.20 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.49 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 116.54 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 131.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 119.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 103.00 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 112.58 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 101.29 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 141.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 129.97 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 103.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 91.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.89 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 116.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 105.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 145.08 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 133.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 118.18 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 106.98 KB
fmhaSm103aKernel_QkvE4m3OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 96.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 85.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 99.09 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 87.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 97.60 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 85.80 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 100.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 88.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.38 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 95.89 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.92 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.73 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.86 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.36 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.85 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 100.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 102.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 97.65 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.27 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 99.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.73 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.53 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 100.50 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.80 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.19 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.39 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.04 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 104.05 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.25 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.45 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.84 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.11 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.70 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.90 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.41 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.56 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 118.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 135.69 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 121.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.85 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 114.76 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.18 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.40 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 111.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.20 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.38 KB
fmhaSm103aKernel_QkvE4m3OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.71 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.03 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 94.11 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.01 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 96.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.65 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 94.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 97.32 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 136.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 123.60 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 138.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 125.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.04 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.31 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 114.33 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.68 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 109.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 137.72 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 124.01 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 140.43 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 126.61 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.69 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 113.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.40 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.81 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 96.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 84.47 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 100.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 88.45 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 97.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 112.56 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 89.00 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 140.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 117.78 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.44 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 90.94 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 85.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 101.21 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 113.10 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 92.71 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 142.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 121.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 115.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 94.77 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 89.13 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.08 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.33 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.24 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.40 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.22 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.37 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.13 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.57 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 108.62 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.20 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.59 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.16 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.48 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.58 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.45 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.60 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.39 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.75 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.50 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.71 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.66 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.70 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.10 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.17 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.09 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.21 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.98 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.33 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 116.35 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.06 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.55 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.68 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.82 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.02 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.30 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.98 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.79 KB
fmhaSm103aKernel_QkvE4m3OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.43 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 112.69 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 89.12 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 140.27 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 117.91 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 114.56 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 90.94 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1PersistentContext.cubin 22-Jul-2026 16:22 113.23 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK16SageV1StaticContext.cubin 22-Jul-2026 16:22 92.84 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1PersistentContext.cubin 22-Jul-2026 16:22 142.15 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK1SageV1StaticContext.cubin 22-Jul-2026 16:22 121.52 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1PersistentContext.cubin 22-Jul-2026 16:22 115.34 KB
fmhaSm103aKernel_QkvE4m3OFp16H64SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128SageQ1SageK4SageV1StaticContext.cubin 22-Jul-2026 16:22 94.77 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.47 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 95.42 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.17 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 99.16 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.16 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.45 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.05 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.24 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 117.16 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.07 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.75 KB
fmhaSm103aKernel_QkvFp16OFp16H128PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.94 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.59 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 118.05 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.28 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 121.72 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.45 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.12 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.12 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 115.80 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.38 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.24 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.07 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.95 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.66 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.96 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.21 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.66 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.22 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 118.70 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.86 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 122.55 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.31 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.91 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.08 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 116.89 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.78 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.18 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.55 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.05 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 120.18 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.62 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.94 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.62 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 140.03 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 129.59 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 143.58 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 133.01 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 134.90 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 123.66 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 138.56 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 127.19 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.84 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 119.79 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 134.38 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 123.34 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 132.27 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 120.81 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 135.94 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 124.41 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 130.88 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 125.59 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 122.30 KB
fmhaSm103aKernel_QkvFp16OFp16H128PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 122.98 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 107.28 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.57 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.18 KB
fmhaSm103aKernel_QkvFp16OFp16H128SeparateQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.24 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.41 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 87.94 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 105.11 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 91.09 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.07 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 90.96 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 105.16 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 93.91 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.72 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 98.54 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.73 KB
fmhaSm103aKernel_QkvFp16OFp16H256PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.63 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 142.80 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 123.43 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 145.73 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 126.22 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.71 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.54 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.46 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 110.33 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.79 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.59 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.66 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 104.38 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.14 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 101.29 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.09 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 104.09 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 143.34 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 128.19 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 146.20 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 131.13 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.89 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.24 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.86 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.19 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.63 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 105.21 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.61 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 108.16 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.28 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 104.00 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.26 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvDensePagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 107.07 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 154.17 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 134.73 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 157.26 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 137.77 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 135.52 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 118.82 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 138.62 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 121.74 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.48 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 112.88 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.55 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.92 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.86 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 113.05 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.81 KB
fmhaSm103aKernel_QkvFp16OFp16H256PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 115.84 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 111.13 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 99.41 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.84 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvCausalVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 102.66 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 110.83 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 102.55 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.80 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PackedQkvDenseVarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 105.51 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 161.59 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 152.18 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 164.68 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 155.09 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 138.86 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 125.87 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 142.08 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 128.78 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 124.95 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 114.47 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.18 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvCausalP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 117.26 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 160.60 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 154.62 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 163.58 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP16VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 157.56 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 140.50 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 129.82 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 143.48 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP32VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 132.77 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.20 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 117.94 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.04 KB
fmhaSm103aKernel_QkvFp16OFp16H512HVPerCta256PagedKvDenseP64VarSeqSkipsSoftmaxQ128Kv128StaticContext.cubin 22-Jul-2026 16:22 120.88 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.14 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 90.34 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.27 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.34 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 102.55 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 91.19 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 109.61 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 98.27 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.80 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 101.45 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 119.71 KB
fmhaSm103aKernel_QkvFp16OFp16H64PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 108.34 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.04 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 106.17 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.66 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.69 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.69 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.90 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.55 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.91 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.14 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.49 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.02 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.38 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.05 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.13 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.04 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.02 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.16 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.02 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.08 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 114.53 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.20 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.74 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.09 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.75 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 114.66 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.22 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.55 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.35 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 115.69 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 104.01 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.45 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvDensePagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.02 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.58 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 117.27 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 136.50 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 124.28 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.10 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 115.00 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 133.38 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 121.49 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.56 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.59 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 131.84 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 119.97 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.36 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 115.29 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 133.78 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingOrChunkedCausalPagedVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 121.66 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP16VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 122.26 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP32VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 119.96 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomP64VarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 118.56 KB
fmhaSm103aKernel_QkvFp16OFp16H64PagedKvSlidingWindowCustomPagedVarSeqQ256Kv128StaticGroupedKeepsAbForGen.cubin 22-Jul-2026 16:22 119.88 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.01 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 89.54 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 112.54 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 96.55 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 106.42 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 90.25 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 113.62 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvDenseVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 97.47 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 116.80 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 100.34 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 123.11 KB
fmhaSm103aKernel_QkvFp16OFp16H80PackedQkvSlidingOrChunkedCausalVarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 107.37 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 127.81 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 134.23 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 118.09 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 122.02 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.81 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.42 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 112.70 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.78 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 102.19 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 125.06 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 109.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 128.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 111.77 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 135.75 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 118.79 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 121.70 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 105.99 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 129.12 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.13 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 118.59 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 103.03 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 126.00 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvDenseP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 110.05 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 139.23 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 122.30 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 145.66 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP16VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 128.69 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 133.56 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 116.92 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 139.86 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP32VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 123.43 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 130.20 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 113.30 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128PersistentContext.cubin 22-Jul-2026 16:22 136.62 KB
fmhaSm103aKernel_QkvFp16OFp16H80PagedKvSlidingOrChunkedCausalP64VarSeqSkipsSoftmaxQ256Kv128StaticContext.cubin 22-Jul-2026 16:22 119.80 KB
LICENSE 22-Jul-2026 16:31 11.09 KB
Artifactory/7.117.25 Server