NVIDIA / NVIDIA/cutlass

[QST] :NCxHWx interleaved conv weight layout uses CxRSKx instead of KCxRSx

Open
#3,010 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

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

Description

Hi CUTLASS team,
I’m looking at the NCxHWx interleaved convolution path and noticed that the weight layout is described/used as CxRSKx rather than KCxRSx.
Intuitively, KxRSCx seems like it would provide more contiguous access for the weight, which could improve memory coalescing and cache behavior. However, CUTLASS uses CxRSKx for this case.

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 locating the NCxHWx interleaved convolution implementation and its weight-layout definitions. Compare the documented and used CxRSKx ordering with the proposed KCxRSx or KxRSCx ordering; done means determining whether the current layout is intentional and resolving the question through the relevant implementation or documentation.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
performance
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.