This is an interesting adventure. Needs OpenShift Needs a compute card or GPU Needs Node Feature Discovery (NFD) Operator Needs Storage Needs Subscription entitlement
Lsantillan passed along this blog post. https://kenmoini.com/blog/using-nvidia-gpus-in-openshift/
Followed this Blog mostly https://www.openshift.com/blog/how-to-use-entitled-image-builds-to-build-drivercontainers-with-ubi-on-openshift
Template borrowed for creating MachineConfig for entitlement https://raw.githubusercontent.com/openshift-psap/blog-artifacts/master/how-to-use-entitled-builds-with-ubi/0003-cluster-wide-machineconfigs.yaml.template
Took parts of this for setting up a MachineConfig to entitle all nodes, didn't use Helm https://docs.nvidia.com/datacenter/kubernetes/pdf/openshift-on-gpu-install-guide.pdf
Used this as reference https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/getting-started.html
Troubleshooting note mentions Entitlement stuff https://docs.nvidia.com/datacenter/kubernetes/openshift-on-gpu-install-guide/index.html#openshift-gpu-troubleshooting
Git location for Operator https://github.com/NVIDIA/gpu-operator