Skip to content

Add CUDA architecture to CMake option when building for NVidia devices #2029

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.

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

Open
ndgrigorian opened this issue Mar 24, 2025 · 0 comments · May be fixed by #2096
Open

Add CUDA architecture to CMake option when building for NVidia devices #2029

ndgrigorian opened this issue Mar 24, 2025 · 0 comments · May be fixed by #2096
Assignees
Labels
build cmake Issue pertaining to CMake enhancement New feature or request good first issue Good for newcomers

Comments

@ndgrigorian
Copy link
Collaborator

Currently, DPCTL_TARGET_CUDA cmake option is binary, and doesn't allow the user to set a CUDA architecture.

This could become problematic in the future and/or for extensions which the compiler may generate code which is unusable on some architectures.

The solution is to use DPCTL_TARGET_CUDA option to allow the user to set an architecture, and if one isn't sett, to fall back to the default (sm_50 is the current default per oneAPI extension)

@ndgrigorian ndgrigorian changed the title Add CUDA architecture parameter when building for NVidia devices Add CUDA architecture to CMake option when building for NVidia devices Mar 24, 2025
@ndgrigorian ndgrigorian added good first issue Good for newcomers enhancement New feature or request build cmake Issue pertaining to CMake labels Mar 24, 2025
@vlad-perevezentsev vlad-perevezentsev self-assigned this Jun 5, 2025
@vlad-perevezentsev vlad-perevezentsev linked a pull request Jun 5, 2025 that will close this issue
8 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
build cmake Issue pertaining to CMake enhancement New feature or request good first issue Good for newcomers
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants