@Morpheus0x Your feature request has been noted and will be put in the backlog.

It appears to be trivial so it should we'll try to do it soon but I cannot tell you when.

@Andrew awesome, thank you for your explanation. After reading a bit more about MTU (see this) I get it now. I looked up MTU in OPNsense for the VLAN interfaces and was surprised that it automatically uses 1496 MTU, which explains why it works and means I don't have to adjust my host's MTU settings, which I really didn't want to do.

Now, as far as I understand, keeping it this way could introduce some split packets and therefore potentially limiting throughput. To fix this, I would need to increase the MTU to 1504 or larger. However, after some more reading (thanks for the keyword "baby giants"), it seems that today, MTU is just a "guideline" for LAN. I will keep the host MTU at 1500 for now.

Please check your XOA is fully up to date 🙂

The only GPU I've tried to pass through has been an old Quadro 2000 card, and it seemed to work fine (though the old driver is a bit difficult on Win Server 2022). But the host also has a cheapo server GPU for XCP-NG to use for boot.

I mention Quadro because those didn't have the crippled firmware blocking them from pass through like the Geforce cards have/had. Kind of a dirty trick from nVidia!

Sorry I don't have any newer cards that I can test with, and I also only have Quadro cards here at all, "required" for some of our software.

What I did find out is that if you pass through a GPU, you can not migrate that VM to another host, even if that other host has the same model GPU installed. And with that my GPU pass though testing is complete.

@toby4213

It would show as many GPU as phisical procesors available.

Also what performance could I expect from one VM with a passed through GPU?
Regarding the use of GPU full , other performace depends of other resurces alocated for the VM. (disk, memory,...)

With passthrough does NVENC work in the VM?
YES

With the use of docker you can do as 20 nvenc per GPU... with K1,, we do 14 max with K2200 (1 GPU visible and passed thru to one machine)

We do not encourage to modify the dom0 Console IP is returned by XAPI. So check your management network IP address.