Skip to content

SM100 Groupwise GeMM K-Major Scale Supports#1102

Merged
yzh119 merged 2 commits intoflashinfer-ai:mainfrom
cyx-6:sm100-gemm-pdl
May 31, 2025
Merged

SM100 Groupwise GeMM K-Major Scale Supports#1102
yzh119 merged 2 commits intoflashinfer-ai:mainfrom
cyx-6:sm100-gemm-pdl

Conversation

@cyx-6
Copy link
Copy Markdown
Collaborator

@cyx-6 cyx-6 commented May 30, 2025

📌 Description

  1. Add support for SM100 groupwise GeMM K-major scale inputs
  2. Add support for switching MMA SM for SM100 groupwise GeMM
  3. Enable SM100 groupwise GeMM PDL(Programmatic Dependent Launch)
  4. Upgrade CUTLASS version to 4.0

🔍 Related Issues

🚀 Pull Request Checklist

Thank you for contributing to FlashInfer! Before we review your pull request, please make sure the following items are complete.

✅ Pre-commit Checks

  • I have installed pre-commit by running pip install pre-commit (or used your preferred method).
  • I have installed the hooks with pre-commit install.
  • I have run the hooks manually with pre-commit run --all-files and fixed any reported issues.

If you are unsure about how to set up pre-commit, see the pre-commit documentation.

🧪 Tests

  • Tests have been added or updated as needed.
  • All tests are passing (unittest, etc.).

Reviewer Notes

Comment thread flashinfer/gemm.py Outdated
The granularity of the scale tensor, (m_granularity, n_granularity, k_granularity).

scale_major_k: bool
Whether scale tensors are k-major layout.
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

better to explain what's k-major (and the alternative mn-major) because it can be confusing to new users.

The name scale_major_k is also confusing, can we usescale_major_mode: str = "MN" or "K" instead?

Copy link
Copy Markdown
Collaborator

@yzh119 yzh119 left a comment

Choose a reason for hiding this comment

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

LGTM

@yzh119 yzh119 merged commit 976cfb4 into flashinfer-ai:main May 31, 2025
2 checks passed
@cyx-6 cyx-6 deleted the sm100-gemm-pdl branch February 3, 2026 08:05
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.

2 participants