Skip to content

FlyDSL gemm_decode: small-M dense GEMM kernels (BF16/FP8/blockscale)#757

Draft
vedenev-amd wants to merge 1 commit into
mainfrom
GEMM_small_M_SILOTIGER-669
Draft

FlyDSL gemm_decode: small-M dense GEMM kernels (BF16/FP8/blockscale)#757
vedenev-amd wants to merge 1 commit into
mainfrom
GEMM_small_M_SILOTIGER-669

2 times slower then Sami's CK C++ krenel

4f37916
Select commit
Loading
Failed to load commit list.