Where to pull amzn2-suffixed driver images from?

I’m attempting to deploy the GPU Operator (23.9.2) on AWS EKS and have a node group configured with instance type p3.8xlarge nodes. When a node is provisioned, the GPU Operator (appropriately) launches pods onto that node due to its node label feature.node.kubernetes.io/pci-10de.present=true which identifies it as an NVIDIA GPU node.

One of the pods launched is the nvidia-driver-daemonset which appears to be responsible for installing the driver, based on the version in the helm charts (550.54.14) and formulates (internally) a driver image name of nvcr.io/nvidia/driver:550.54.14-amzn2.

However, the POD soon enters into an ImagePullBackOff loop due to this error:

Failed to pull image "nvcr.io/nvidia/driver:550.54.14-amzn2": rpc error: code = NotFound desc = failed to pull and unpack image "nvcr.io/nvidia/driver:550.54.14-amzn2": failed to resolve reference "nvcr.io/nvidia/driver:550.54.14-amzn2": nvcr.io/nvidia/driver:550.54.14-amzn2: not found

When consulting the NCG Catalog for its list of drivers (tags), nowhere is there a tag suffixed with -amzn2. Since this particular portion of the tag is applied by the GPU Operator itself, how do we go about resolving these kinds of images?

Given the popularity of AWS EKS as a provider and the fact that GPU Operator has been around for some time, I’m a little concerned I’m missing something fairly obvious since I’m unable to find references to this issue - which should be reasonably common.

Thanks for any help that can be provided.