Sharing GPU between host and guest

Question

Is there a way to make both host and guest use a singular GPU without any paid software (vGPU)?

Bonus points if it:

Dynamically allocates GPU resources
Works for AMD and Intel GPUS

I have an nVidia GPU, a Linux host and a Windows guest.

Perhaps a Linux alternative to https://github.com/jamesstringerparsec/Easy-GPU-PV

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. — Community, Apr 07 '23 at 12:29
@Peregrino69 passthrough is an exclusive thing, not a shared thing. — Marcus Müller, Apr 07 '23 at 13:24
Thx @MarcusMüller :-) I was searching that as OP tagged gpu-passthrough. — Peregrino69, Apr 07 '23 at 13:28

score 0 · Answer 1 · answered Apr 07 '23 at 13:23

Both spice, used by Qemu/KVM, as well as virtualbox, offer accelerated 3D from the guest, which essentially means they translate openGL calls.

That is sharing the host GPU with the VM, through the well-standardized openGL API.

If you want to, however, do something like using GPGPU (e.g CUDA, openCL):

No, that is currently not possible. If you want to virtualize parts of your GPU, you will need a datacenter GPU from nvidia, and the datacenter license for CUDA. I'm not aware of any of that being available with other vendors.

I found this, but seems to be only for windows: https://github.com/jamesstringerparsec/Easy-GPU-PV — Anm, Apr 07 '23 at 15:45

Sharing GPU between host and guest

1 Answers1