ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-15 18:42:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CK_TILE] Refine fp8 support in flatmm (#2239)

Browse Source 
* [CK_TILE] Refine fp8 in flatmm

1. Replace USING_MFMA_16x16x32 & USING_MFMA_16x16x32 with constexpr
2. Add an additional const check to avoid build error in HotLoopScheduler
3. Refine shuffleb to support both tile 32x32 and 16x16
4. Support command option -init
5. Move Gemm warp defintion to a separate struct

* fix clang format

* fix clang format

* keep default bhavior unchanged (warp tile = 16x16)

* fix tile engine build error

* fix a typo in codegen_utils.py

* address review comments

* address review comments

---------

Co-authored-by: Thomas Ning <Thomas.Ning@amd.com>

[ROCm/composable_kernel commit: 37e1a27537]
This commit is contained in:
linqunAMD
2025-06-25 16:07:45 +08:00
committed by GitHub
parent b62e551ccb
commit d2ec53a74e
10 changed files with 313 additions and 198 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tile_engine/ops/gemm/codegen_utils.py
View File

@@ -44,9 +44,12 @@ CSHUFFLE_EPILOGUE = """
using GemmEpilogue = ck_tile::CShuffleEpilogue<
ck_tile::CShuffleEpilogueProblem<ADataType,
BDataType,
ck_tile::tuple<>,
AccDataType,
CDataType,
ck_tile::tuple<>,
CLayout,
ck_tile::element_wise::PassThrough,
GemmPipelineProblem::kBlockSize,
TilePartitioner::MPerBlock,
TilePartitioner::NPerBlock,