NVIDIA / NVIDIA/cutlass

[QST]How to write column major output epilogue in cutlass2.x

Open
#2,258 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

? - Needs Triage inactive-30d inactive-90d question
Dominant language
C++
Stars
10.5k
Forks
2.1k
Avg merge
3d 11h
Merged PRs (30d)
7

Description

Hello, I implement fp8 sparse gemm on sm89 with the 15th example and it works correctly. However, I just find that the column major output not implemented yet, as the former issue said. Could you please tell how to write a column major output epilogue ?

@hwu36 @thakkarV

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 with CUTLASS 2.x example 15 and the former issue #1150 to understand the reported epilogue limitation. Trace the existing fp8 sparse GEMM output epilogue and determine what is required for column-major output on SM89. Done means the requested column-major output path is implemented and verified for that example.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.