Skip to content

Apply moe_reduce_sum kernel for fused_marlin_moe#12888

Merged
ispobock merged 3 commits intomainfrom
ke/use-moe-reduce-sum
Nov 8, 2025
Merged

Apply moe_reduce_sum kernel for fused_marlin_moe#12888
ispobock merged 3 commits intomainfrom
ke/use-moe-reduce-sum

Conversation

@ispobock
Copy link
Copy Markdown
Collaborator

@ispobock ispobock commented Nov 8, 2025 •
edited
Loading

Motivation

before apply (bs1): 4.6μs
image
after apply (bs1): 1.5μs
image

@github-actions github-actions Bot added sgl-kernel and removed run-ci labels Nov 8, 2025
@gemini-code-assist
Copy link
Copy Markdown
Contributor

gemini-code-assist Bot commented Nov 8, 2025

Summary of Changes

Hello @ispobock, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly enhances the performance of the fused_marlin_moe operation by integrating a custom, optimized sum reduction kernel. This change streamlines the computation of Mixture of Experts layers, leading to faster execution times and improved efficiency, particularly for inference workloads.

Highlights

  • Performance Optimization: Implemented a specialized moe_reduce_sum kernel within the fused_marlin_moe function, resulting in a significant performance improvement (e.g., 4.6us down to 1.5us for batch size 1).
  • Kernel Integration: Replaced the generic torch.sum operation with a call to the newly imported moe_sum_reduce kernel for efficient sum reduction.
  • New Parameter: Introduced routed_scaling_factor as a new parameter to the fused_marlin_moe and fused_marlin_moe_fake functions, allowing for dynamic scaling of routed outputs.
Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature Command Description
Code Review /gemini review Performs a code review for the current pull request in its current state.
Pull Request Summary /gemini summary Provides a summary of the current pull request in its current state.
Comment @gemini-code-assist Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help /gemini help Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

  1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist[bot]
gemini-code-assist Bot reviewed Nov 8, 2025
View reviewed changes
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

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

Code Review

This pull request introduces a significant performance optimization for the fused_marlin_moe function by replacing torch.sum with a custom moe_sum_reduce CUDA kernel. This change, as indicated in the description, provides a substantial speedup. The implementation is clean and also introduces a routed_scaling_factor for added flexibility. The function signatures have been consistently updated across real and fake functions. I have one minor suggestion to improve code clarity.

Comment thread sgl-kernel/python/sgl_kernel/fused_moe.py Outdated
@ispobock ispobock mentioned this pull request Nov 8, 2025
Closed
9 tasks
@ispobock ispobock added the run-ci label Nov 8, 2025
@ispobock @gemini-code-assist
Update sgl-kernel/python/sgl_kernel/fused_moe.py
4d98f26 
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
@ispobock ispobock merged commit 44f594d into main Nov 8, 2025
25 of 57 checks passed
@ispobock ispobock deleted the ke/use-moe-reduce-sum branch November 8, 2025 17:31
@ispobock ispobock mentioned this pull request Dec 10, 2025
Merged
6 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@zhyncs zhyncs Awaiting requested review from zhyncs

@BBuf BBuf Awaiting requested review from BBuf BBuf is a code owner

@yizhang2077 yizhang2077 Awaiting requested review from yizhang2077 yizhang2077 is a code owner

@merrymercy merrymercy Awaiting requested review from merrymercy merrymercy is a code owner

@FlamingoPg FlamingoPg Awaiting requested review from FlamingoPg FlamingoPg is a code owner

@HaiShaw HaiShaw Awaiting requested review from HaiShaw HaiShaw is a code owner

1 more reviewer

@gemini-code-assist gemini-code-assist[bot] gemini-code-assist[bot] left review comments

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

run-ci sgl-kernel

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@ispobock @sglang-bot