Tesla M40 as second GPU Passthrough

[sentein]

Right now i have a 60TB array with 2 xeon 2650v2 256gb of ram and with a plex docker using an P600 for transcoding. I have a few other dockers running but that is the only one with any real hardware associated with it. I am using the Nvidia plugin on version 6.9.2. With all that out of the way, i am attempting to use either a Tesla M40 or a Tesla K80 passed through to a VM as a second GPU. The main idea behind this is i want all of my rendering horse power on the server so i can design in fusion 360 or use blender from almost any machine i own. I would like to use the original VNC as the base GPU so i can have easy access to the VM anywhere in the house.

 

As of right now i have tried to set the Tesla M40 up as the second GPU in the VM but Windows does not see the hardware at all. I do not have a Vbios for it either. I have also tried to use the card as just a PCIE device but unraid will not let me pass it through that way. I would rather not setup a second machine if i can help it. Is there something i can do to get either of these cards to pass through correctly? I have enough system resources to choke a horse on my server i would rather keep it in one box.

[loadedlouie]

Hi did you managed to put it to work ? 

 

I have a M40 24Gb and i'm trying to set it up the vm see's it, but appers with a error code...

thanks in advance 

 

[mikeyosm]

DId this work? I'm thinking of buying a Telsa P4 and splitting the vgpu across multiple Windows 11 VMs

[MacModMachine]

On 10/12/2022 at 8:22 AM, mikeyosm said:

DId this work? I'm thinking of buying a Telsa P4 and splitting the vgpu across multiple Windows 11 VMs

 

you can use vGPU , its a complicated process and very hacky. i have 2 p4's split up myself. 

Edited January 20, 2023 by MacModMachine

[lincolnliu]

@MacModMachine can you share some pointers or tutorials on how to setup the P4 in unraid as a vgpu?