Skip to content

[fix] WAR to fix the illegal memory access issue in moe gemm on SM120 #5636

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

Jump to bottom
Merged
kaiyux merged 3 commits into from
Jul 10, 2025
Merged

[fix] WAR to fix the illegal memory access issue in moe gemm on SM120 #5636

kaiyux merged 3 commits into from
Jul 10, 2025

Conversation

@peaceh-nv
Copy link
Collaborator

@peaceh-nv peaceh-nv commented Jul 1, 2025 •
edited
Loading

Description

Avoid using void elementc to fix the illegal memory access issue in moe gemm on SM120 after cutlass upgrade and add related unit tests on SM120

@peaceh-nv
Copy link
Collaborator Author

peaceh-nv commented Jul 1, 2025

/bot run

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 1, 2025

PR_Github #10450 [ run ] triggered by Bot

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 1, 2025

PR_Github #10450 [ run ] completed with state FAILURE
/LLM/main/L0_MergeRequest_PR pipeline #7732 completed with status: 'FAILURE'

@peaceh-nv
Copy link
Collaborator Author

peaceh-nv commented Jul 1, 2025

/bot run

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 1, 2025

PR_Github #10496 [ run ] triggered by Bot

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 1, 2025

PR_Github #10496 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #7771 completed with status: 'FAILURE'

@pamelap-nvidia
Copy link
Collaborator

pamelap-nvidia commented Jul 1, 2025

/bot run --disable-fail-fast

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 1, 2025

PR_Github #10530 [ run ] triggered by Bot

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 1, 2025

PR_Github #10530 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #7794 completed with status: 'FAILURE'

@peaceh-nv peaceh-nv changed the title [fix] Downgrade cutlass to fix the illegal memory access issue in moe gemm on SM120 [fix] WAR to fix the illegal memory access issue in moe gemm on SM120 Jul 8, 2025
@peaceh-nv peaceh-nv requested a review from djns99 July 8, 2025 09:25
@peaceh-nv
Copy link
Collaborator Author

peaceh-nv commented Jul 9, 2025

/bot run

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 9, 2025

PR_Github #11424 [ run ] triggered by Bot

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 9, 2025

PR_Github #11424 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8448 completed with status: 'FAILURE'

@peaceh-nv
Copy link
Collaborator Author

peaceh-nv commented Jul 9, 2025

/bot run

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 9, 2025

PR_Github #11437 [ run ] triggered by Bot

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 9, 2025

PR_Github #11437 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8459 completed with status: 'SUCCESS'

@kaiyux kaiyux merged commit 76c3a12 into NVIDIA:main Jul 10, 2025
3 checks passed
zhou-yuxin pushed a commit to zhou-yuxin/TensorRT-LLM that referenced this pull request Jul 15, 2025
@peaceh-nv @zhou-yuxin
[fix] WAR to fix the illegal memory access issue in moe gemm on SM120 (…
f545152 
…NVIDIA#5636)

Signed-off-by: peaceh <[email protected]>
Signed-off-by: Yuxin <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@djns99 djns99 djns99 approved these changes

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

5 participants

@peaceh-nv @tensorrt-cicd @pamelap-nvidia @djns99 @kaiyux