MiniMax-AI / MiniMax-AI/MSA

Feature Request: Support for SM90 (Hopper / H100) GPUs

Open
#13 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
422
Forks
57
Avg merge
15h 50m
Merged PRs (30d)
1

Description

Summary

Is there any plan to support SM90 (Hopper / H100) GPUs? Currently MSA only supports NVIDIA SM100 (compute capability 10.0), but many users have access to Hopper-based GPUs like H100/H800 which use compute capability 9.0 (sm_90).

Details

  • Target GPU architecture: SM90 / Hopper (e.g., H100 SXM, H100 PCIe, H800)
  • Compute capability: 9.0 (sm_90)
  • Current status: Not mentioned in README, no related issues found, code directory only has fmha_sm100/

Questions

  1. Is SM90 support on the roadmap?
  2. If so, any estimated timeline?
  3. If not, would the team accept community contributions for SM90 porting?

Thanks!

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reading the README and inspecting the code directory, especially fmha_sm100/. Compare the existing SM100 support with the requirements for SM90/Hopper GPUs. Done would mean defining and completing the scope of SM90 support, or documenting a clear roadmap and contribution path.

Written by the indexing model from the issue text.

Assessment

Domain
hpc
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.