Skip to content

[TRTLLM-6222][feat] Extend cute_dsl_nvfp4_gemm to sm103. #9543

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
kaiyux merged 3 commits into from
Dec 1, 2025
Merged

[TRTLLM-6222][feat] Extend cute_dsl_nvfp4_gemm to sm103. #9543

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
c6651fc
extend cute_dsl_nvfp4_gemm to sm103.
limin2021 Nov 28, 2025
4c63efa
extend cute_dsl_nvfp4_gemm to sm103.
limin2021 Nov 28, 2025
28f892f
update doc.
limin2021 Dec 1, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -69,7 +69,7 @@ triton==3.5.0; platform_machine == "x86_64"
tiktoken
blobfile
openai-harmony==0.0.4
nvidia-cutlass-dsl==4.3.0; python_version >= "3.10"
nvidia-cutlass-dsl==4.3.1; python_version >= "3.10"
Comment thread
kaiyux marked this conversation as resolved.
plotly
numexpr<2.14.0 # WAR for attempted use of nonexistent numpy.typing
partial_json_parser
4 changes: 2 additions & 2 deletions tensorrt_llm/_torch/custom_ops/cute_dsl_custom_ops.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,9 +49,9 @@ def __init__(self, alpha: float, output_dtype: torch.dtype):
self.output_dtype = output_dtype
assert output_dtype == torch.bfloat16

if get_sm_version() != 100:
if get_sm_version() not in [100, 103]:
raise ValueError(
f"SM version {get_sm_version()} is not supported for {self.__class__.__name__}, it only supports SM 100"
f"SM version {get_sm_version()} is not supported for {self.__class__.__name__}, it only supports SM 100 and SM 103"
)

# rewrite the hash function because the value of self.alpha doesn't affect the tactic.
Expand Down
4 changes: 2 additions & 2 deletions tests/unittest/_torch/thop/parallel/test_fp4_linear.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,8 +87,8 @@ def test_fp4_linear(dtype, mnk):
@pytest.mark.skipif(sys.version_info < (3, 12),
reason="cutlass-dsl 4.1.0 requires Python 3.12+")
@pytest.mark.skipif(
get_sm_version() != 100,
reason="This test is only supported in Blackwell architecture",
get_sm_version() not in [100, 103],
reason="This test is only supported in sm100 and sm103 architecture",
)
@pytest.mark.skipif(not IS_CUTLASS_DSL_AVAILABLE,
reason="cutlass-dsl is not available")
Expand Down
Loading