NVIDIA / NVIDIA/cutlass

[QST]How Does TMA Work in CUTLASS for Writing from Shared Memory to Global Memory?

Open
#2,008 5 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

Could you explain how TMA works in CUTLASS? For example, when writing from the shared memory Tensor sS to the global memory Tensor gD, it seems that the data is written sequentially, i.e., sS[i] directly maps to gD[i]. Is this the correct behavior?

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 tracing the CUTLASS TMA operation for the shared-memory tensor sS and global-memory tensor gD. A useful response should explain whether sS[i] maps directly to gD[i] and describe the addressing or layout behavior that determines the write order.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
hpc
Issue type
Documentation
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.