pytorch / pytorch/audio

No support for Blackwell GPUs on cu128

Open
#3,892 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
2.9k
Forks
799
Avg merge
58m
Merged PRs (30d)
3

Description

🚀 The feature

Trying to use nightly pyTorch builds which use cu128 on Blackwell NVIDIA GPUs.

"torchvision 0.22.0.dev20250316+cu128 requires torch==2.8.0.dev20250315+cu128, but you have torch 2.6.0 which is incompatible."

Motivation, pitch

NVIDIA's new GPUs use new cu128, so things like torchaudio/torchvision should be updated to 2.8.0 along with various other things.

Alternatives

No response

Additional context

No response

Contributor guide

Open the contributing guide

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 nightly PyTorch builds using cu128, Blackwell NVIDIA GPUs, and version compatibility between torch, torchvision, and torchaudio, but no files or tests. Start by locating the repository's dependency and build configuration, then determine the required version updates and verify that the cu128 combinations are compatible.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.