[bugfix]: require CUDA 12.9 for SM103a VSA builds - #1814
Open
SolitaryThinker wants to merge 1 commit into
Open
Conversation
Keep CUDA 12.8 GB200 builds from receiving the unsupported compute_103a gencode. Export whether the extension actually contains an SM103a image so runtime dispatch falls back safely when a CUDA 12.8 build is moved to B300.
mergify
Bot
added
type: bugfix
Contributor
Merge Protections🔴 1 of 1 protections blocking · waiting on 👀 reviews and 🤖 CI
🔴 PR merge requirementsWaiting for
This rule is failing.
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
This follow-up to #1812 preserves CUDA 12.8 support for GB200 builds.
CUDA 12.8 supports SM100a, while CUDA 12.9 introduced the SM103a compiler target. The previous CMake configuration added
compute_103awhenever native VSA was enabled, which caused CUDA 12.8 builds targeting SM100a to fail.Changes
Testing
python -m pytest -q tests/test_fasth3_packaging.py— 3 passed.pre-commit run --files fastvideo-kernel/CMakeLists.txt fastvideo-kernel/csrc/common_extension.cpp fastvideo-kernel/python/fastvideo_kernel/block_sparse_attn_sm100a.py tests/test_fasth3_packaging.py— passed for all applicable hooks.A full CUDA build was not run because this host does not provide a C++ compiler or
nvcc. CI should validate the CUDA 12.8 and CUDA 13 build configurations.Impact
CUDA 13 release wheels continue to include SM100a and SM103a cubins. CUDA 12.8 builds retain SM100a support. Training and gradient-enabled VSA paths continue to use Triton and are unchanged.