wheels: drop sm_75 SASS to fit EB under PyPI 320 MB limit#293
Merged
Conversation
Fatbin compression had no effect (auditwheel --strip likely strips the compressed sections). Drop the Turing (sm_75) SASS target instead — each SASS architecture is a full copy of device code (~50 MB). T4 users JIT-compile from the sm_90 PTX on first kernel launch (~10-30s one-time cost). Keep sm_80-real (A100/A10) as the primary SASS target. AMReX_CUDA_ARCH and CMAKE_CUDA_ARCHITECTURES both updated to match: 80-real;90-virtual (was 75-real;80-real;90-virtual).
Code Coverage ReportGenerated by CI — coverage data from gcovr |
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fatbin compression had no effect (auditwheel --strip likely strips the compressed sections). Drop the Turing (sm_75) SASS target instead — each SASS architecture is a full copy of device code (~50 MB). T4 users JIT-compile from the sm_90 PTX on first kernel launch (~10-30s one-time cost). Keep sm_80-real (A100/A10) as the primary SASS target.
AMReX_CUDA_ARCH and CMAKE_CUDA_ARCHITECTURES both updated to match: 80-real;90-virtual (was 75-real;80-real;90-virtual).