Thanks to visit codestin.com
Credit goes to github.com

Skip to content

cuda: Use common gfx8 value for GCN4 #11209

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

Closed
wants to merge 1 commit into from

Conversation

Haus1
Copy link
Contributor

@Haus1 Haus1 commented Jan 12, 2025

rocm_agent_enumerator returns the two values 800 and 803 with the former being used by this API

Make sure to read the contributing guidelines before submitting a PR

rocm_agent_enumerator returns the two values 800 and 803 with the former being used by this API
@github-actions github-actions bot added Nvidia GPU Issues specific to Nvidia GPUs ggml changes relating to the ggml tensor library for machine learning labels Jan 12, 2025
@arch-btw
Copy link
Contributor

Here is my output:

-> rocm_agent_enumerator
gfx000
gfx803
gfx902

Will 803 not be functional anymore?

@Haus1
Copy link
Contributor Author

Haus1 commented Jan 13, 2025

Here is my output:

-> rocm_agent_enumerator
gfx000
gfx803
gfx902

Will 803 not be functional anymore?

It only returns one ID per device although ROCM can have multiple, you'll need to check to see which one is actually being used. The gfx000 should say gfx800, which version of ROCM are you using?

edit:
it seems this may be a regression in ROCm: ROCm/ROCm#223

@Haus1
Copy link
Contributor Author

Haus1 commented Jan 13, 2025

While looking at the code to parse the device properties it appears the minor version is shifted up a decimal place such that any device IDs requiring the first digit will never match. I'm going to hold back on this for a moment since it requires more attention than I initially thought.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
ggml changes relating to the ggml tensor library for machine learning Nvidia GPU Issues specific to Nvidia GPUs
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants