Is there anything we need to do for simulating GPUs?

[drmorr0]

We can simulate a node that's running GPUs right now, just by including the GPU resource in the simulated node spec. However, GPUs are complicated in terms of actually running workloads, especially with the new (DRA)[https://github.com/NVIDIA/k8s-dra-driver/blob/main/README.md] stuff that's been happening. We'll probably need to do some back-end work to support this somehow, maybe both in the tracer and the simulated environment?