[CUDA] Adjust package build settings for Blackwell GPU #26307
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
1 participant
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Outdated suggestions cannot be applied.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.
Description
We recently upgraded CI pipeline to CUDA 12.8 (See #26267), which allows us to build native code for Blackwell GPU of compute capability 120. Here we update the CMAKE_CUDA_ARCHITECTURES settings.
Cloud GPUs GB200, GB300 etc have compute capability of 100 and 103, which are excluded due to package size limit.
Since CUDA 12.4, there is a new nvcc option for compress mode:
https://docs.nvidia.com/cuda/cuda-compiler-driver-nvcc/#compress-mode-default-size-speed-balance-none-compress-mode
Note that this option is not compatible with drivers released before CUDA 12.4 Release so we only apply it to CUDA 13 build. That could reduce package size for cuda 13 significantly.
Python wheel size:
Motivation and Context
Binary size can be reduced as reported in #26282