wait_for_manager_pods_deployed in tests/virt/node/gpu/utils.py actually waits for a DaemonSet to be deployed specifically in the nvidia-gpu-operator namespace. A caller has to read the implementation to discover this.
Rename to something like wait_for_nvidia_gpu_operator_ds_deployed (or wait_for_nvidia_ds_ready) so the NVIDIA/DaemonSet scope is clear from the name alone.