If this is your first visit, be sure to
check out the FAQ by clicking the
link above. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite
Seems interesting. PCI/VGA passthrough has been around for a long time but requires an IOMMU unit on the motherboard and CPU, BIOS, GPU support for VT-d or AMD-V for anything to work.
So this is virtualizing the GPU rather than passing it directly into the VM? One problem with VGA passthrough is that by passing it over, you prevent the host OS from accessing the GPU hardware. Thus its necessary to have two GPUs from my understanding.
Michael, can you tell us if this technology would work with Windows guests inside KVM and support hardware accelerated Direct3D? Or is this only a Linux guest technology?
- Basic functions of full GPU virtualization works, guest can see a full-featured vGPU.
We ran several 3D workloads such as lightsmark, nexuiz, urbanterror and warsow.
So yes, it's really a vgpu, and this can be shared across multivm. I think for the display output you need a remote protocol like spice or vnc.
Does this work without VT-d? Second it will be fine if I could use my integrated gpu for VMs and on the host system my main Nvidia card on standard PC not laptop.
Comment