NVIDIA / NVIDIA/cutlass

[QST] FP4 Tensor Core mma.sync Instruction Unsupported on SM_101 Architecture

Open
#2,908 7 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

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

Description

The current kernel uses the following FP4 Tensor Core instruction:
mma.sync.aligned.m16n8k64.row.col.kind::mxf4nvf4.block_scale.scale_vec::4X.f32.e2m1.e2m1.f32.ue4m3

This instruction is only supported on SM_120+ (Hopper/Blackwell architectures). Our target platform is SM_101, which does not have native FP4 Tensor Core support, so the instruction cannot execute.

Is there a recommended method to emulate this FP4 Tensor Core MMA on SM_101 while maintaining equivalent numerical results?

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 kernel that emits mma.sync.aligned.m16n8k64.row.col.kind::mxf4nvf4.block_scale.scale_vec::4X.f32.e2m1.e2m1.f32.ue4m3 and checking the SM_101 and SM_120 capability paths. No files or tests are named in the issue, so identify the relevant CUTLASS kernel and validation coverage first. Done means an SM_101-compatible emulation exists and produces equivalent numerical results.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
backend
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.