NVIDIA / NVIDIA/cutlass

[QST] thread num assert in sm70_epilogue_vectorized

Open
#1,334 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

Hi, @thakkarV
https://github.com/NVIDIA/cutlass/blob/47a3ebbea9860e14c095b52c4e6e2db33340f572/include/cutlass/epilogue/collective/sm70_epilogue_vectorized.hpp#L237

Strangely, it requires TiledCopyS2R's threads equal to the MMA AtomC's threads. I think here we describe how each thread does LDS and therefore it should be:

CUTE_STATIC_ASSERT(typename TiledCopyS2R::TiledNumThr{} == thr_size(typename TiledMma{}));

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 at include/cutlass/epilogue/collective/sm70_epilogue_vectorized.hpp around line 237 and read the surrounding TiledCopyS2R and TiledMma assertions. Verify which thread-count relationship the epilogue requires, then validate the change with the relevant CUTLASS build or tests. Done means the assertion expresses the intended relationship without breaking supported configurations.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
hpc
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.