Add warning about removed sm50 and sm60 arches #158301
Closed
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
atalman
added
topic: not user facing
atalman
commented
Jul 15, 2025
torch/cuda/__init__.py
Outdated
| ) | ||
| if current_arch < min_arch: | ||
| warnings.warn( | ||
| old_gpu_warn |
There was a problem hiding this comment.
Looks like incorrect_binary_warn is never used. However its probably more accurate warning
nWEIdia
approved these changes
Jul 15, 2025
|
Wait, these warning are saying two opposite things lol. Can we rationalize these messages to be more aligned with the state of the world:
|
|
@pytorchmergebot merge -f "lint is green" |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
Just adding a note that in future, we might want to re-evaluate for "cur_arch > max_arch" case, as there could be scenarios that the binary may be able to still support. But "cur_arch < min_arch" is definitely not supported. e.g. suppose we build sm up to sm80, running on sm86 would still work. Similarly along this line for sm120. |
|
@pytorchbot cherry-pick --onto release/2.8 -c critical |
pytorchbot
pushed a commit
that referenced
this pull request
Jul 16, 2025
Related to #157517 Detect when users are executing torch build with cuda 12.8/12.9 and running on Maxwell or Pascal architectures. We would like to include reference to the issue: #157517 as well as ask people to install CUDA 12.6 builds if they are running on sm50 or sm60 architectures. Test: ``` >>> torch.cuda.get_arch_list() ['sm_70', 'sm_75', 'sm_80', 'sm_86', 'sm_90', 'sm_100', 'sm_120', 'compute_120'] >>> torch.cuda.init() /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:263: UserWarning: Found <GPU Name> which is of cuda capability 5.0. PyTorch no longer supports this GPU because it is too old. The minimum cuda capability supported by this library is 7.0. warnings.warn( /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:268: UserWarning: Support for Maxwell and Pascal architectures is removed for CUDA 12.8+ builds. Please see #157517 Please install CUDA 12.6 builds if you require Maxwell or Pascal support. ``` Pull Request resolved: #158301 Approved by: https://github.com/nWEIdia, https://github.com/albanD (cherry picked from commit fb731fe)
Cherry picking #158301The cherry pick PR is at #158478 and it is recommended to link a critical cherry pick PR with an issue. The following tracker issues are updated: Details for Dev Infra teamRaised by workflow job |
atalman
added a commit
that referenced
this pull request
Jul 16, 2025
Add warning about removed sm50 and sm60 arches (#158301) Related to #157517 Detect when users are executing torch build with cuda 12.8/12.9 and running on Maxwell or Pascal architectures. We would like to include reference to the issue: #157517 as well as ask people to install CUDA 12.6 builds if they are running on sm50 or sm60 architectures. Test: ``` >>> torch.cuda.get_arch_list() ['sm_70', 'sm_75', 'sm_80', 'sm_86', 'sm_90', 'sm_100', 'sm_120', 'compute_120'] >>> torch.cuda.init() /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:263: UserWarning: Found <GPU Name> which is of cuda capability 5.0. PyTorch no longer supports this GPU because it is too old. The minimum cuda capability supported by this library is 7.0. warnings.warn( /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:268: UserWarning: Support for Maxwell and Pascal architectures is removed for CUDA 12.8+ builds. Please see #157517 Please install CUDA 12.6 builds if you require Maxwell or Pascal support. ``` Pull Request resolved: #158301 Approved by: https://github.com/nWEIdia, https://github.com/albanD (cherry picked from commit fb731fe) Co-authored-by: atalman <atalman@fb.com>
|
@pytorchbot revert -m="Diff reverted internally" -c="ghfirst" This Pull Request has been reverted by a revert inside Meta. To re-land this change, please open another pull request, assign the same reviewers, fix the CI failures that caused the revert and make sure that the failing CI runs on the PR by applying the proper ciflow label (e.g., ciflow/trunk).) |
|
@pytorchbot successfully started a revert job. Check the current status here. |
pytorchmergebot
added a commit
that referenced
this pull request
Jul 19, 2025
This reverts commit fb731fe. Reverted #158301 on behalf of https://github.com/facebook-github-bot due to Diff reverted internally ([comment](#158301 (comment)))
|
@atalman your PR has been successfully reverted. |
atalman
force-pushed
the
add_warning_about_old_sm
branch
from
e17ea11 to
7706737
Compare
Move code fixes Revert "conda" This reverts commit 2853662. Revert "use tos accept" This reverts commit 8b34264. Revert "conda" This reverts commit 2853662. Revert "Revert "conda"" This reverts commit e732654. Revert "Revert "use tos accept"" This reverts commit c456c54. Revert "Revert "conda"" This reverts commit bb4fa09. fix fix_arch_list fix fixes
atalman
force-pushed
the
add_warning_about_old_sm
branch
from
e8cd442 to
100002a
Compare
atalman
commented
Jul 19, 2025
| if torch.version.cuda is not None: # on ROCm we don't want this check | ||
| CUDA_VERSION = torch._C._cuda_getCompiledVersion() # noqa: F841 | ||
| if ( | ||
| torch.version.cuda is not None and torch.cuda.get_arch_list() |
There was a problem hiding this comment.
New version added check for torch.cuda.get_arch_list()
pytorchmergebot
pushed a commit
that referenced
this pull request
Jul 20, 2025
Related to #157517 Detect when users are executing torch build with cuda 12.8/12.9 and running on Maxwell or Pascal architectures. We would like to include reference to the issue: #157517 as well as ask people to install CUDA 12.6 builds if they are running on sm50 or sm60 architectures. Test: ``` >>> torch.cuda.get_arch_list() ['sm_70', 'sm_75', 'sm_80', 'sm_86', 'sm_90', 'sm_100', 'sm_120', 'compute_120'] >>> torch.cuda.init() /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:263: UserWarning: Found <GPU Name> which is of cuda capability 5.0. PyTorch no longer supports this GPU because it is too old. The minimum cuda capability supported by this library is 7.0. warnings.warn( /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:268: UserWarning: Support for Maxwell and Pascal architectures is removed for CUDA 12.8+ builds. Please see #157517 Please install CUDA 12.6 builds if you require Maxwell or Pascal support. ``` Please note I reverted original PR #158301 because it broke internal users. This is a reland, added added check for non empty torch.cuda.get_arch_list() Pull Request resolved: #158700 Approved by: https://github.com/huydhn, https://github.com/Skylion007, https://github.com/eqy
tvukovic-amd
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Aug 20, 2025
Add warning about removed sm50 and sm60 arches (pytorch#158301) Related to pytorch#157517 Detect when users are executing torch build with cuda 12.8/12.9 and running on Maxwell or Pascal architectures. We would like to include reference to the issue: pytorch#157517 as well as ask people to install CUDA 12.6 builds if they are running on sm50 or sm60 architectures. Test: ``` >>> torch.cuda.get_arch_list() ['sm_70', 'sm_75', 'sm_80', 'sm_86', 'sm_90', 'sm_100', 'sm_120', 'compute_120'] >>> torch.cuda.init() /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:263: UserWarning: Found <GPU Name> which is of cuda capability 5.0. PyTorch no longer supports this GPU because it is too old. The minimum cuda capability supported by this library is 7.0. warnings.warn( /home/atalman/.conda/envs/py312/lib/python3.12/site-packages/torch/cuda/__init__.py:268: UserWarning: Support for Maxwell and Pascal architectures is removed for CUDA 12.8+ builds. Please see pytorch#157517 Please install CUDA 12.6 builds if you require Maxwell or Pascal support. ``` Pull Request resolved: pytorch#158301 Approved by: https://github.com/nWEIdia, https://github.com/albanD (cherry picked from commit fb731fe) Co-authored-by: atalman <atalman@fb.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 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.
Related to #157517
Detect when users are executing torch build with cuda 12.8/12.9 and running on Maxwell or Pascal architectures.
We would like to include reference to the issue: #157517 as well as ask people to install CUDA 12.6 builds if they are running on sm50 or sm60 architectures.
Test:
cc @ptrblck @msaroufim @eqy @jerryzh168 @albanD @malfet