Skip to content

[Cherry-Pick][BugFix] cherry-pick #7238 support moe for sm103#7240

Merged
zoooo0820 merged 1 commit into
PaddlePaddle:release/2.6from
BingooYang:2.6/moe_sm103
Apr 8, 2026
Merged

[Cherry-Pick][BugFix] cherry-pick #7238 support moe for sm103#7240
zoooo0820 merged 1 commit into
PaddlePaddle:release/2.6from
BingooYang:2.6/moe_sm103

Conversation

@BingooYang
Copy link
Copy Markdown
Contributor

@BingooYang BingooYang commented Apr 8, 2026

Motivation

支持moe sm103编译

Modifications

支持sm103编译

Usage or Command

no

Accuracy Tests

no

Checklist

  • Add at least a tag in the PR title.
    • Tag list: [[FDConfig],[APIServer],[Engine], [Scheduler], [PD Disaggregation], [Executor], [Graph Optimization], [Speculative Decoding], [RL], [Models], [Quantization], [Loader], [OP], [KVCache], [DataProcessor], [BugFix], [Docs], [CI], [Optimization], [Feature], [Benchmark], [Others], [XPU], [HPU], [GCU], [DCU], [Iluvatar], [Metax]]
    • You can add new tags based on the PR content, but the semantics must be clear.
  • Format your code, run pre-commit before commit.
  • Add unit tests. Please write the reason in this PR if no unit tests.
  • Provide accuracy results.
  • If the current PR is submitting to the release branch, make sure the PR has been submitted to the develop branch, then cherry-pick it to the release branch with the [Cherry-Pick] PR tag.

@paddle-bot
Copy link
Copy Markdown

paddle-bot Bot commented Apr 8, 2026

Thanks for your contribution!

Copy link
Copy Markdown

@PaddlePaddle-bot PaddlePaddle-bot left a comment

Choose a reason for hiding this comment

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

🤖 AI Code Review | 2026-04-08

📋 Review 摘要

PR 概述:支持 MOE GEMM kernel 在 SM103 (Compute Capability 10.3) 设备上编译运行

变更范围custom_ops/gpu_ops/cutlass_kernels/moe_gemm/ (2 个文件)

影响面 Tag[BugFix] [OP]

📝 PR 规范检查

✅ 标题包含有效 Tag [BugFix]
✅ 描述包含 Motivation 和 Modifications 部分

问题

级别 文件 概述
- - -

未发现阻塞性问题。

总体评价

变更简单直接,通过扩展架构检查范围使 SM103 设备能够编译 MOE kernel。SM103 (Compute Capability 10.3) 将使用 Sm80 kernel 运行,这在没有专用 kernel 实现的情况下是合理的方案。

技术细节

  1. 编译时检查__CUDA_ARCH__ < 1010< 1100,使 SM103 (1030) 在 Sm80 kernel 编译范围内
  2. 运行时检查sm_ < 101< 104,同步修改运行时 SM 版本分发逻辑
  3. 影响范围:SM80-SM103 (8.0-10.3) 所有设备都使用 Sm80 kernel

@codecov-commenter
Copy link
Copy Markdown

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (release/2.6@7ab48c4). Learn more about missing BASE report.

Additional details and impacted files
@@              Coverage Diff               @@
##             release/2.6    #7240   +/-   ##
==============================================
  Coverage               ?   73.87%           
==============================================
  Files                  ?      376           
  Lines                  ?    52886           
  Branches               ?     8249           
==============================================
  Hits                   ?    39071           
  Misses                 ?    11089           
  Partials               ?     2726           
Flag Coverage Δ
GPU 73.87% <ø> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@BingooYang BingooYang changed the title [BugFix] support moe for sm103 [Cherry-Pick][BugFix] support moe for sm103 Apr 8, 2026
@zoooo0820 zoooo0820 changed the title [Cherry-Pick][BugFix] support moe for sm103 [Cherry-Pick][BugFix] cherry-pick #7238 support moe for sm103 Apr 8, 2026
@zoooo0820 zoooo0820 merged commit 0181884 into PaddlePaddle:release/2.6 Apr 8, 2026
68 of 75 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants