Closed asmacdo closed 8 months ago
Enable GPU Profiles: https://github.com/dandi/dandi-hub/blob/do-eks/helm/jupyterhub/dandihub.yaml#L90-L134
We will also need to enable karpenter resources to accommodate these profiles: https://github.com/dandi/dandi-hub/blob/do-eks/addons.tf#L170-L259
Can we also do matlab in this PR or should that be separated?
T4 GPU for inference 8 CPU / 30 GB / 1 T4 GPU Base (MATLAB) 6 CPU / 16 GB upto 12C/32G T4 GPU for inference (MATLAB) 8 CPU / 30 GB / 1 T4 GPU.
Additionally, we need to bound the upper limit of usage, especially for the gpu nodes.
Done via 7812e04fbc212e29cacf97f3964798a9767dd83e
Enable GPU Profiles: https://github.com/dandi/dandi-hub/blob/do-eks/helm/jupyterhub/dandihub.yaml#L90-L134
We will also need to enable karpenter resources to accommodate these profiles: https://github.com/dandi/dandi-hub/blob/do-eks/addons.tf#L170-L259
Can we also do matlab in this PR or should that be separated?
T4 GPU for inference 8 CPU / 30 GB / 1 T4 GPU Base (MATLAB) 6 CPU / 16 GB upto 12C/32G T4 GPU for inference (MATLAB) 8 CPU / 30 GB / 1 T4 GPU.
Additionally, we need to bound the upper limit of usage, especially for the gpu nodes.