Theoretical dedicated gpu vm that paravirtualizes to other guest VMs

[mustang-landmine]

I was just wondering is it theoretically possible to have a pcie dedicated GPU in a vm that then paravirtualizes its graphics to another guest vm. Or is this not how paravirtualization works

Like this:
PCIE GPU VM —> virgl maybe → app qube

Also is it it safer to do PCIE gpu vm to paravirtualization app qube, or is there no good reason to do this and just do straight up pcie gpu passthrough to the app qube

Has something like this existed or am I missing something

[Atrate]

There’s a different approach that works with newer CPUs:

Virtualised Intel GPU with SR-IOV

A Guide to virtualise your Intel GPU. Heavily based on Lenovo ThinkPad T14 Gen 5 - #27 by tze. The Output will still be send trough the normal QubesOS pipeline, so no extra Monitor is needed. ATTENTION: This Guide requires executing 3rd party Code in dom0 at Kernel priveleges (a driver). This can be a security risk depending on your threadmodell, but you also can check the code (if you understand a bit of programming). All commands must be run in dom0. You need this Patchset from Github. Pi…

[Atrate]

Also on the topic of a dedicated GPU VM:

GUI domain