How would you describe them?
Paravirtual NVIDIA GPU for KVM guests — run unmodified CUDA, PyTorch and Vulkan inside a VM at host parity, on a GPU the host keeps using. No passthrough, no vGPU licence. Experimental.