Skip to content

Add gpu_type and use correct value for GPU accounting#18

Open
ZQyou wants to merge 1 commit into
mainfrom
update_for_gpu_accounting
Open

Add gpu_type and use correct value for GPU accounting#18
ZQyou wants to merge 1 commit into
mainfrom
update_for_gpu_accounting

Conversation

@ZQyou

@ZQyou ZQyou commented Jul 1, 2026

Copy link
Copy Markdown

No description provided.

@ZQyou ZQyou requested a review from treydock July 1, 2026 20:34
Comment thread submit.yml.erb
node_selector:
node-role.kubernetes.io/ondemand: ''
nvidia.com/gpu.product: <%= extra_gpu_mem.to_i.positive? ? 'NVIDIA-A100-PCIE-40GB-MIG-7g.40gb' : 'NVIDIA-A100-PCIE-40GB-MIG-3g.20gb' %>
gpu_type: <%= extra_gpu_mem.to_i.positive? ? 'nvidia.com/mig-7g.40gb' : 'nvidia.com/mig-3g.20gb' %>

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

We no longer have 20GB MIG nodes. We have only 40GB and 5GB at this time. The 5GB would be nvidia.com/mig-1g.5gb I think.

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

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

@ejaffe-osc told me that the default GB option may break most users' workflows. What happened to the 20GB option?

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

It was never used and we needed more 40GB devices so the 20GB option was removed.

@ZQyou ZQyou self-assigned this Jul 1, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants