Skip to content

[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8 fmha kernels on Ada. #5779

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
merged 2 commits into from
Jul 8, 2025
Merged

[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8 fmha kernels on Ada. #5779

merged 2 commits into from
Jul 8, 2025

Conversation

@PerkzZheng
Copy link
Collaborator

@PerkzZheng PerkzZheng commented Jul 7, 2025

Fallback to cubins for now, and revisit it later.

qsang-nv and others added 2 commits July 7, 2025 02:56
@qsang-nv @PerkzZheng
keep sm90 headsize 128 cubins (NVIDIA#5320)
cb57544 
Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
@PerkzZheng
fallback to cubins for ada fp8 fmha kernels
a445ab2 
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
@PerkzZheng PerkzZheng requested a review from a team as a code owner July 7, 2025 02:59
@PerkzZheng
Copy link
Collaborator Author

PerkzZheng commented Jul 7, 2025

/bot run

@PerkzZheng PerkzZheng requested review from QiJune and qsang-nv July 7, 2025 02:59
@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 7, 2025

PR_Github #11092 [ run ] triggered by Bot

@tensorrt-cicd
Copy link
Collaborator

tensorrt-cicd commented Jul 7, 2025

PR_Github #11092 [ run ] completed with state SUCCESS
/LLM/release-0.21/L0_MergeRequest_PR pipeline #176 completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

@crazydemo
Copy link
Collaborator

crazydemo commented Jul 8, 2025

verified the fp8 cases, all pass.

@crazydemo crazydemo self-requested a review July 8, 2025 02:19
@QiJune QiJune requested a review from litaotju July 8, 2025 02:26
@litaotju litaotju merged commit 5a50e2b into NVIDIA:release/0.21 Jul 8, 2025
4 checks passed
dc3671 pushed a commit to dc3671/TensorRT-LLM that referenced this pull request Jul 10, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
f2c22e4 
… fmha kernels on Ada. (NVIDIA#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
dc3671 pushed a commit to dc3671/TensorRT-LLM that referenced this pull request Jul 10, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
23709b6 
… fmha kernels on Ada. (NVIDIA#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
dc3671 pushed a commit to dc3671/TensorRT-LLM that referenced this pull request Jul 10, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
1699401 
… fmha kernels on Ada. (NVIDIA#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
dc3671 pushed a commit to dc3671/TensorRT-LLM that referenced this pull request Jul 11, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
a82503f 
… fmha kernels on Ada. (NVIDIA#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
dc3671 pushed a commit to dc3671/TensorRT-LLM that referenced this pull request Jul 14, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
b2d9838 
… fmha kernels on Ada. (NVIDIA#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
dc3671 pushed a commit to dc3671/TensorRT-LLM that referenced this pull request Jul 14, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
b87f5be 
… fmha kernels on Ada. (NVIDIA#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
dc3671 pushed a commit to dc3671/TensorRT-LLM that referenced this pull request Jul 14, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
a644be5 
… fmha kernels on Ada. (NVIDIA#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
dc3671 pushed a commit that referenced this pull request Jul 14, 2025
@PerkzZheng @qsang-nv @dc3671
[https://nvbugspro.nvidia.com/bug/5355054] fallback to cubins for fp8…
4a0b7a0 
… fmha kernels on Ada. (#5779)

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
Co-authored-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@litaotju litaotju litaotju approved these changes

@QiJune QiJune QiJune approved these changes

@crazydemo crazydemo crazydemo approved these changes

+1 more reviewer

@qsang-nv qsang-nv qsang-nv approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

6 participants

@PerkzZheng @tensorrt-cicd @crazydemo @litaotju @QiJune @qsang-nv