NVIDIA / NVIDIA/cutlass

[QST]Are Tensors Equivalent After Different Layout Transformations?

Open
#1,777 2 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

What is your question?
I first have an array of pointers A. Then, following layout B, I obtain a tensor T0. Similarly, I have another array of pointers B (different from A), and by following layout C, I obtain a tensor T1. After printing, I see that T0 and T1 are the same. If we then proceed with subsequent operations, such as thread_mma.partition_A, will the results be the same? Can T0 and T1 be considered equivalent in this context?

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

The issue names layout B, layout C, tensors T0 and T1, and thread_mma.partition_A, but no files or tests. Start by locating the CUTLASS documentation or examples for tensor layouts and partition_A. Done means providing an authoritative explanation of whether equal printed tensors remain equivalent for subsequent operations.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
hpc
Issue type
Documentation
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.