[Perf] SM90 cutlass fp8 mm supports odd M by swap_ab, 180~290% kernel performance improvement (#44572)

Signed-off-by: yewentao256 <zhyanwentao@126.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
This commit is contained in:
Wentao Ye
2026-06-13 12:05:45 -07:00
committed by GitHub
co-authored by mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
parent 521b88c29e
commit 71b961dd35
3 changed files with 87 additions and 159 deletions
@@ -245,8 +245,6 @@ def test_cutlass_fp8_blockwise_scale_gemm(
return
if m % a_scale_group_shape[0] != 0 or k % a_scale_group_shape[1] != 0:
return
if m % 4 != 0 and current_platform.has_device_capability(100):
return
cutlass_fp8_gemm_helper(m, n, k, a_scale_group_shape, b_scale_group_shape, use_bias)