Jump to content

Floppy012

Member
  • Posts

    1
  • Joined

  • Last visited

Awards

This user doesn't have any awards

Floppy012's Achievements

  1. Hey, about a year ago after I've seen this video from LTT I built myself a new PC with a single GPU I already had lying around. I managed to setup the PCI Passtorugh. And it ran flawlessly. But since I'm mainly using Linux I noticed over time, that my CPU wasn't that good in powering 4 FHD displays. So I disabled the passthrough and used my dedicated GPU for powering my Linux OS. This is my current setup: MB: ASRock Z370 Extreme 4 CPU: Intel i7 8700K GPU: GTX 1050 TI (currently powering linux) (PCIE2) GPU2: GT710 (PCIE6) SSD: 250GB NVMe (M2_1) And 4 FHD Displays Now I want to upgrade my system to allow me to work on Linux and play games like Rainbow Six Siege on Windows. I borrowed a GPU from a friend but sadly the two GPUs are in the same IOMMU Group which doesn't allow me to Passthrough only one of them into the VM. Also in my UEFI I can only choose between Internal Graphics and GPU for startup (even with both GPUs installed). But as you see in my setup list I currently have two GPUs installed. This because if I put a GPU into PCIE6 both GPUs are in different IOMMU Groups. So I currently have the GT710 in passthrough to access my Windows machine. Sadly PCIE6 only supports x4 if another GPU is installed. I now thought, that I could buy myself a second GPU, put it into the PCIE4 slot and then split the IOMMU Group somehow (I read about splitting IOMMU groups but can't remember where), which would allow me to passthrough only one of the GPUs. But before I buy some expensive GPU I wanted to ask here if there is anyone who has a better idea/experience of how to accomplish this. TL;DR: I want to have two useful GPUs in my system and passthrough one of them into a Windows KVM Thanks in advance Edit: I forgot to mention, that all displays need to be powered by my Linux GPU but not necessarily by my Windows GPU (two would be plenty)
×