feat(torch): Support compute capability 12.0 #99
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Compute Capability 12.0 + CUDA 12.9 Fixes
This change adds compute capability 12.0 (
sm_120) to the list of compiled architectures for libraries in theml-containers/torchandml-containers/torch-extrasimages. It additionally fixes an issue with CUDA 12.9 support when compiling PyTorch extension modules by cherry-picking a patch from themainbranch of the PyTorch repo into ourtorchbuild, and getstorchto use the system NVTX library when available (i.e. on CUDA 12.9).xformersinml-containers/torch-extrashas also been updated to its v0.0.30 release, offering better compatibility with newer GPU architectures, PyTorch versions, and with vLLM.Finally, this update drops compute capability 7.0 (e.g. V100), leaving 7.5 (e.g. Quadro RTX 4000/5000) as the lowest supported architecture.