Dev/a8w4 and a8w8splitk (#3447)

* Ck moe bs splitk pr (#3440)

* splitk kick-off. Compilation fail

* splitk hack pass

* fix scale offset calc.

* clang-format for a8w8_moe_blk_gemm1 splitk change

* fix testcase error

---------

Co-authored-by: oscar <huaiguxu@amd.com>
Co-authored-by: huaiguxu <145733371+huaiguxu@users.noreply.github.com>

* Zan/moe a8w4 (#3441)

* update

* update

* update ck moe a8w4

* update

* update

* update

* compile pass

* update

* update

* python3 op_tests/test_moe_2stage.py -t 16 -e 1 -k 1 -dim 256,256 ready

* support new a8w4 kernel

* update

* update ck_tile

* re format

* update

* update

* fix conflict

* fix build

* update ck_tile moe

* fix clang format

* fix the problem

* fix accruacy issue

* fix

---------

Co-authored-by: oscar <huaiguxu@amd.com>
Co-authored-by: huaiguxu <145733371+huaiguxu@users.noreply.github.com>
Co-authored-by: Zzz9990 <zanzhang@amd.com>
Co-authored-by: felix <felix.li@amd.com>
This commit is contained in:
yadaish
2025-12-19 09:26:52 +08:00
committed by GitHub
parent ba897f8435
commit c0ee71d735
13 changed files with 2911 additions and 139 deletions

View File

@@ -309,25 +309,6 @@ using WarpGemmMfma_f32_16x16x64_bf8_bf8 = WarpGemmImpl<WarpGemmAttributeMfmaIter
template <typename A, typename B, WGAttrNumAccessEnum AttrNumAccess = WGAttrNumAccessEnum::Single>
using WarpGemmMfma_f32_16x16x128_f8f6f4 = WarpGemmImpl<
WarpGemmAttributeMfma<WarpGemmAttributeMfmaImpl_f32_16x16x128_f8f6f4<A, B>, AttrNumAccess>>;
template <WGAttrNumAccessEnum AttrNumAccess = WGAttrNumAccessEnum::Single>
using WarpGemmMfma_f32_16x16x128_fp8_fp8 = WarpGemmImpl< //
WarpGemmAttributeMfma<WarpGemmAttributeMfmaImpl_f32_16x16x128_f8f6f4<fp8_t, fp8_t>,
AttrNumAccess>>;
template <WGAttrNumAccessEnum AttrNumAccess = WGAttrNumAccessEnum::Single>
using WarpGemmMfma_f32_16x16x128_fp8_bf8 = WarpGemmImpl< //
WarpGemmAttributeMfma<WarpGemmAttributeMfmaImpl_f32_16x16x128_f8f6f4<fp8_t, bf8_t>,
AttrNumAccess>>;
template <WGAttrNumAccessEnum AttrNumAccess = WGAttrNumAccessEnum::Single>
using WarpGemmMfma_f32_16x16x128_bf8_fp8 = WarpGemmImpl< //
WarpGemmAttributeMfma<WarpGemmAttributeMfmaImpl_f32_16x16x128_f8f6f4<bf8_t, fp8_t>,
AttrNumAccess>>;
template <WGAttrNumAccessEnum AttrNumAccess = WGAttrNumAccessEnum::Single>
using WarpGemmMfma_f32_16x16x128_bf8_bf8 = WarpGemmImpl< //
WarpGemmAttributeMfma<WarpGemmAttributeMfmaImpl_f32_16x16x128_f8f6f4<bf8_t, bf8_t>,
AttrNumAccess>>;
template <WGAttrNumAccessEnum AttrNumAccess = WGAttrNumAccessEnum::Single>
using WarpGemmMfma_f32_16x16x128_fp8_fp8_CTransposed =