[lovedswain]

Running certain ML models in VMs

Running CUDA in VMs

Running transcoders in VMs

Running <anything that needs a GPU> in VMs

[ur-whale]

This is the exact same information you posted above.

Please see my edit.