Skip to content

[Kernel] Add fp8_w8a8 fused MoE kernel tuning configs for DeepSeek V3/R1 on NVIDIA H20 #16753

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
Apr 17, 2025
Merged

[Kernel] Add fp8_w8a8 fused MoE kernel tuning configs for DeepSeek V3/R1 on NVIDIA H20 #16753

merged 2 commits into from
Apr 17, 2025

Conversation

Ximingwang-09
Copy link
Contributor

@Ximingwang-09 Ximingwang-09 commented Apr 17, 2025 •
edited
Loading

Refer to sgl-project/sglang#5196, add fp8_w8a8 fused MoE kernel tuning configs for DeepSeek V3/R1 on H20. There is a significant throughput improvements compared to default configurations.

@github-actions GitHub Actions
Copy link

👋 Hi! Thank you for contributing to the vLLM project.

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

@Ximingwang-09 Ximingwang-09 mentioned this pull request Apr 17, 2025
Merged
ximing.wxm added 2 commits April 17, 2025 12:55
@jeejeelee jeejeelee requested a review from simon-mo April 17, 2025 06:52
mgoin
mgoin approved these changes Apr 17, 2025
View reviewed changes
Copy link
Member

@mgoin mgoin left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks!

@github-project-automation github-project-automation bot moved this from Backlog to In progress in DeepSeek V3/R1 Apr 17, 2025
@mgoin mgoin added the ready ONLY add when PR is ready to merge/full CI is needed label Apr 17, 2025
@DarkLight1337 DarkLight1337 merged commit a018e55 into vllm-project:main Apr 17, 2025
60 checks passed
@github-project-automation github-project-automation bot moved this from In progress to Done in DeepSeek V3/R1 Apr 17, 2025
@Ximingwang-09 Ximingwang-09 deleted the tuned_config branch April 18, 2025 02:09
yangw-dev pushed a commit to yangw-dev/vllm that referenced this pull request Apr 21, 2025
@Ximingwang-09 @yangw-dev
[Kernel] Add fp8_w8a8 fused MoE kernel tuning configs for DeepSeek V3…
a9d5e0f 
…/R1 on NVIDIA H20 (vllm-project#16753)

Signed-off-by: ximing.wxm <[email protected]>
Co-authored-by: ximing.wxm <[email protected]>
Signed-off-by: Yang Wang <[email protected]>
@ayrnb
Copy link

ayrnb commented Apr 27, 2025

H20 96G or 141G？

@Ximingwang-09
Copy link
Contributor Author

H20 96G or 141G？

96G

jikunshang pushed a commit to jikunshang/vllm that referenced this pull request Apr 29, 2025
@Ximingwang-09 @dbyoung18
[Kernel] Add fp8_w8a8 fused MoE kernel tuning configs for DeepSeek V3…
cb49f7b 
…/R1 on NVIDIA H20 (vllm-project#16753)

Signed-off-by: ximing.wxm <[email protected]>
Co-authored-by: ximing.wxm <[email protected]>
@handsome-chips
Copy link

Is there a configuration for the H800?

lk-chen pushed a commit to lk-chen/vllm that referenced this pull request Apr 29, 2025
@Ximingwang-09 @lk-chen
[Kernel] Add fp8_w8a8 fused MoE kernel tuning configs for DeepSeek V3…
3473d18 
…/R1 on NVIDIA H20 (vllm-project#16753)

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

@mgoin mgoin mgoin approved these changes

@simon-mo simon-mo Awaiting requested review from simon-mo

Assignees
No one assigned
Labels
ready ONLY add when PR is ready to merge/full CI is needed
Projects
DeepSeek V3/R1
Status: Done
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
5 participants
@Ximingwang-09 @ayrnb @handsome-chips @mgoin @DarkLight1337