[CMake] Fix SM70/72 support range upper bound for CUDA 13.0#3166
Open
Flink-ddd wants to merge 1 commit intoNVIDIA:mainfrom
Open
[CMake] Fix SM70/72 support range upper bound for CUDA 13.0#3166Flink-ddd wants to merge 1 commit intoNVIDIA:mainfrom
Flink-ddd wants to merge 1 commit intoNVIDIA:mainfrom
Conversation
Author
|
Hi @hwu36 , Could you please take a look when you have time? Thanks! |
Collaborator
|
We haven't decided to drop volta. Since some are still use it. Could we only drop it with 13.x compiler? |
Author
|
Sorry for the confusion in the description, SM70/72 are only excluded when CUDA_VERSION >= 13.0, they remain supported for all earlier versions. |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Purpose
Fixes #3164.
The CUDA 13.0 compiler no longer accepts SM70/72 as target architectures, causing a fatal error when CUTLASS_NVCC_ARCHS_SUPPORTED includes them. The existing condition only checked VERSION_GREATER_EQUAL 11.4 with no upper bound.
Fix: add VERSION_LESS 13.0 upper bound so SM70/72 are only included when the compiler supports them. Volta remains fully supported on CUDA < 13.0.
Test Plan
Verified CUTLASS_NVCC_ARCHS_SUPPORTED output with CUDA_VERSION=13.0 before and after the fix.
Test Result
Before fix:
-- CUTLASS_NVCC_ARCHS_SUPPORTED = 70;72;75;80;86;87;89;90;90a;100;...
After fix:
-- CUTLASS_NVCC_ARCHS_SUPPORTED = 75;80;86;87;89;90;90a;100;...