ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-06-06 07:51:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

generalized bpreshuffle pipeline optimization

Browse Source
This commit is contained in:
aska-0096
2025-04-27 11:50:30 +00:00
parent 49338edb1b
commit bc9c819aa4
2 changed files with 148 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
example/65_gemm_multiply_multiply/moe_gemm2_xdl_fp8.cpp
View File

@@ -132,6 +132,7 @@ static constexpr ck::index_t CShuffleMLane = BLOCKSIZE / CShuffleNLane;
static constexpr ck::index_t AK1 = 16 / sizeof(A0DataType);
static constexpr ck::index_t BK1 = 16 / sizeof(B0DataType);
static constexpr ck::index_t EVec = 2;
// TODO: Epilogue performance issue. AtomicAdd lose 15~20% performance compare with Set.
static constexpr ck::index_t D0Vec = 1;
static constexpr ck::index_t D1Vec = 1;
static constexpr ck::index_t D2Vec = 1;