Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Integration of a CI workflow for testing GPU and CPU functionality #21

Open
Tracked by #8
alexfreudenberg opened this issue Jun 9, 2023 · 2 comments · May be fixed by #16
Open
Tracked by #8

Integration of a CI workflow for testing GPU and CPU functionality #21

alexfreudenberg opened this issue Jun 9, 2023 · 2 comments · May be fixed by #16

Comments

@alexfreudenberg
Copy link
Owner

alexfreudenberg commented Jun 9, 2023 •
edited
Loading

Would be nice to have a CI pipeline for building and testing - currently, this fails because the new version of the Intel oneAPI doesn't ship the icc compiler anymore and there is no way to install an old version in the Actions workflow.
Also, the Intel's standalone installer seems to be untested - at least there is no way of pushing this install script beyond the extraction phase.
@alexfreudenberg alexfreudenberg mentioned this issue Jun 9, 2023
Open
5 tasks
@alexfreudenberg alexfreudenberg linked a pull request Jun 9, 2023 that will close this issue
Add GitHub action for automatically testing the build process of the core library #16
Draft
@alexfreudenberg alexfreudenberg changed the title GitHub Actions for automatic testing (possible with GPUs?) GitHub Actions for automatic testing Jun 9, 2023
@alexfreudenberg
Copy link
Owner Author

For testing CUDA code, the runner would need to connect to a GPU cloud instance. Here is a nice reference for a safe implementation of that via ssh: https://docs.gitlab.com/ee/ci/ssh_keys/

@alexfreudenberg alexfreudenberg changed the title GitHub Actions for automatic testing Integration of a CI workflow for testing GPU and CPU functionality Jul 5, 2023
@alexfreudenberg
Copy link
Owner Author

After some research, the most reasonable path toward a CI pipeline seems to set up a separate workflow which is manually triggered and launches a GPU cloud instance. Then, the tagged GPU testing workflow is assigned to a self-hosted runner spawned on this instance.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
1 participant
@alexfreudenberg