GVT-d GPU passthrough VM configuration

Oct 10, 2022
48
2
13
While experimenting with GPU passthrough I could not configure to VM in Proxmox, as instructed by this guide from Intel, I'm having trouble figuring out how to properly configure the `-device` in an existing VM in Proxmox.

According to their instructions, the guest VM should be configured as:

Bash:
#! /bin/bash -x
    /usr/bin/qemu-system-x86_64 \
    -enable-kvm \
    -m 2048 \
    -smp 2 \
    -cpu host \
    -hda /home/img/win10-64.qcow2 \
    -vga none -nographic \
    -device vfio-pci,host=00:02.0,x-igd-gms=2,id=hostdev0,bus=pci.0,addr=0x2,x-igd-opregion=on \
    -usb -usbdevice tablet \
    -vnc :2 \
    -net nic,macaddr=00:16:3e:60:0a:50,addr=0x3 -net tap,script=/etc/qemu-ifup \
    -daemonize
 
While experimenting with GPU passthrough I could not configure to VM in Proxmox, as instructed by this guide from Intel, I'm having trouble figuring out how to properly configure the `-device` in an existing VM in Proxmox.
I have no experience with this, but for mediated Intel passthrough I would follow the Proxmox guide instead: https://pve.proxmox.com/wiki/PCI(e)_Passthrough#_mediated_devices_vgpu_gvt_g

EDIT: Sorry, ignore this post as I got GVT-g and GVT-d confused.
 
Last edited:
@leesteken is it possible to output to the HDMI port using the GVT-g method?
Not as far as I know. I though you wanted GVT-g but I see now that the Intel guide is about GVT-d. Sorry, I got them confused again.
Lots of threads on this forum about PCIe passthrough of Intel integrated graphics (and all the work-arounds that are sometimes needed) but I have no experience with that (because Ivy Bridge at the time did not work at all).