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.

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

Add ada architecture label for compute capability 8.9 #1060

Merged
merged 1 commit into from
Dec 6, 2024

Conversation

elezar
Copy link
Member

@elezar elezar commented Nov 28, 2024

No description provided.

@elezar elezar self-assigned this Nov 28, 2024
if computeMinor < 9 {
return "ampere"
}
return "ada"
Copy link
Contributor

Choose a reason for hiding this comment

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

Should this be ada-lovelace? I saw your comment elsewhere on the CUDA samples using just Ada but I'm not sure if this is the correct source of truth to go by.

Copy link
Member Author

Choose a reason for hiding this comment

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

Copy link
Contributor

Choose a reason for hiding this comment

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

OK. I don't feel too strongly - it mostly just feels inconsistent. Especially since its shortened to "Ada" and not "Lovelace" since all others are the last name, not the first name.

Copy link
Member Author

Choose a reason for hiding this comment

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

Updated to ada-lovelace and created NVIDIA/go-nvlib#49

@elezar elezar force-pushed the add-ada-lovelace-label branch from 72dba88 to 03f25d4 Compare December 3, 2024 20:31
@elezar elezar requested a review from klueska December 4, 2024 11:31
@klueska
Copy link
Contributor

klueska commented Dec 4, 2024

Looks like this needs a rebase

@klueska
Copy link
Contributor

klueska commented Dec 4, 2024

Side question -- should getArchFamily() just be a transformation of nvlib.GetArchitectureAsString()? Is there any particular reason to reply directly on compute capabilities to generate this string?

@elezar elezar force-pushed the add-ada-lovelace-label branch from 03f25d4 to aaccdfd Compare December 6, 2024 10:23
Copy link

copy-pr-bot bot commented Dec 6, 2024

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

@elezar
Copy link
Member Author

elezar commented Dec 6, 2024

Side question -- should getArchFamily() just be a transformation of nvlib.GetArchitectureAsString()? Is there any particular reason to reply directly on compute capabilities to generate this string?

In theory, yes. The issue is that we do support CUDA devices and we would need a bit of refactoring to use that directly. I have some thoughts on how to do that, but didn't want to address it in this PR.

@elezar
Copy link
Member Author

elezar commented Dec 6, 2024

@klueska I have rebased the changes. PTAL.

@elezar elezar merged commit 44b263c into NVIDIA:main Dec 6, 2024
7 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants