Skip to content

Drop CUDA 10.2 - #8287

Merged
esc merged 3 commits into
numba:mainfrom
gmarkall:drop-cuda-10-2
Sep 7, 2022
Merged

esc merged 3 commits into
numba:mainfrom
gmarkall:drop-cuda-10-2

Conversation

@gmarkall

Copy link
Copy Markdown
Member

By the time of the next release (0.57, early 2023) CUDA 10.2 will be more than three years old. Supporting it will require extra maintenance overhead for changes planned in the near term, so it is dropped to reduce this burden.

Notes:

  • I removed a comment about cudadevrt from the Cooperative Groups docs. This comment was only relevant when there was the possibility of cudadevrt being missing for an installation, when (broken) older CUDA toolkit packages that happened to be missing it were still supported. Now they are unsupported, cudadevrt can be assumed to be present so there is no need to explicitly mention it in the CG docs.
  • Previously there was a concern about dropping support for Jetsons by dropping CUDA 10.2 - however, Jetpack 5 (presently in developer preview, on the roadmap for release, prior to the Numba 0.57 release) has CUDA 11.4 so this would no longer be an issue.
  • Some code to handle pre-11.0 toolkits has been simplified / removed.

This comment was only relevant when there was the possibility of
cudadevrt being missing for an installation, when (broken) older CUDA
toolkit packages that happened to be missing it were still supported.
Now they are unsupported, cudadevrt can be assumed to be present so
there is no need to explicitly mention it in the CG docs.
@gmarkall gmarkall added 2 - In Progress CUDA CUDA related issue/PR Effort - short Short size effort needed labels Jul 27, 2022
By the time of the next release (0.57, early 2023) CUDA 10.2 will be
more than three years old. Supporting it will require extra maintenance
overhead for changes planned in the near term, so it is dropped to
reduce this burden.
@gmarkall

Copy link
Copy Markdown
Member Author

gpuci run tests

@gmarkall

Copy link
Copy Markdown
Member Author

(The fail is the current parfors fail that is present on master).

@sklam sklam added this to the Numba 0.57 RC milestone Aug 2, 2022
@sklam sklam self-assigned this Aug 2, 2022

@sklam sklam left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The patch looks good.

Before it's merged, the numba buildfarm needs to be adjusted. We should waiting until 0.56.1 release is done before changing the farm

@sklam sklam added Pending BuildFarm For PRs that have been reviewed but pending a push through our buildfarm and removed 3 - Ready for Review labels Aug 22, 2022
@gmarkall gmarkall added the 4 - Waiting on CI Review etc done, waiting for CI to finish label Aug 23, 2022
@esc

esc commented Sep 2, 2022 •

Copy link
Copy Markdown
Member

Build numba_smoketest_cuda_yaml_155 has started

@gmarkall

gmarkall commented Sep 6, 2022

Copy link
Copy Markdown
Member Author

gpuci run tests

@esc

esc commented Sep 6, 2022

Copy link
Copy Markdown
Member

a9c095e82 has passed on the anaconda internal build farm for all CUDA versions except 10.2 as expected.

@esc esc added BuildFarm Passed For PRs that have been through the buildfarm and passed 5 - Ready to merge Review and testing done, is ready to merge and removed 4 - Waiting on CI Review etc done, waiting for CI to finish Pending BuildFarm For PRs that have been reviewed but pending a push through our buildfarm labels Sep 6, 2022
@esc

esc commented Sep 7, 2022

Copy link
Copy Markdown
Member

Thank you for the patch!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

5 - Ready to merge Review and testing done, is ready to merge BuildFarm Passed For PRs that have been through the buildfarm and passed CUDA CUDA related issue/PR Effort - short Short size effort needed

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants