-
Epic
-
Resolution: Done
-
High
-
None
-
None
-
None
Epic Goal
- Ability to deploy a pod requiring nvidia gpu in AWS
Why is this important?
- This is required for AI/ML workload using peer-pods
Scenarios
- As an OpenShift administrator, I want to configure the NVIDIA GPU instance types that are allowed to be used for workloads deployed as OpenShift sandboxed containers (OSC) in AWS
- As an OpenShift user, I want to deploy my GPU workload using the OSC peer-pods approach in AWS
Acceptance Criteria
(The Epic is complete when...)
- Support for creating pod VM image with NVIDIA driver
- Creating a POD using GPU instance type
- ..
Additional context:
Detailed description about the enhancement - https://docs.google.com/document/d/1u-PGV5vpp2IwwbAzUkmS8hAFDuN6K8yRi0UJOJyOSgc/edit