Hello, I want to get proper 3d acceleration in my Qubes. I plan to achieve this by doing by using NVIDIA vGPU. Here is the plan:
- Get an NVIDIA GPU capable of vGPU
- Create a sys-gpu HVM
- Pass the NVIDIA GPU to sys-gpu
- Create a vGPU in the sys-gpu
- pass the mdev (vgpu device) to another Qube (How???)
- Profit
I know that I could just use a SR-IOV capable GPU but I would also need to run it on the host I assume.
I haven’t purchased the GPU yet BTW, if anyone has an example setup similar to what I’m trying to do on ANY GPU vendor let me know.
But for the time being, the question is, how can I pass an mdev device in one HVM to another HVM (or even PVH?)