NVIDIA/cutlass
1
0
Fork 0
mirror of https://github.com/NVIDIA/cutlass.git synced 2026-04-20 06:48:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Minor fix for testing of blockscaled dense GEMM with TMA prefetch (#2930)

Browse Source 
* new example with TMA prefetch feature targeting for DRAM latency bound cases

* minor fix to resitrct as 100a arch

* typo

* apply arch for whole pytest

---------

Co-authored-by: Questa Wang <questaw@computelab-frontend-7.nvidia.com>
Co-authored-by: Questa Wang <questaw@umbriel-b200-145.ipp4a1.colossus.nvidia.com>
This commit is contained in:
questa-quan-wang
2026-01-05 16:36:03 +08:00
committed by GitHub
parent 3d9de19bb7
commit 2aee73922c
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
test/examples/CuTeDSL/sm_100a/test_dense_blockscaled_gemm_persistent_prefetch.py
View File

@@ -47,7 +47,7 @@ from blackwell.dense_blockscaled_gemm_persistent_prefetch import (
)
import cutlass
pytestmark = [pytest.mark.arch(["100a"])]
@pytest.mark.invalid_case(
lambda: not Sm100BlockScaledPersistentDenseGemmKernel.can_implement(