Nvidia Quadro GV100

Hello all,

Sorry if this is not the right venue for this, but I search the forum and the internet and could not find the information that I am looking for.

I am in the process of purchasing a decommissioned Nvidia GV100 that I am planning on using in a Proxmox (Debian - KVM) host running multiple Windows VM to perform scientific simulations (FEA code which has been optimised for CUDA according to the software vendor)). I have already done GPU passthrough on a handful of cards without too much hassle. My question is around the GV100 and whether this card will behave just like a regular Quadro.

From reading the literature on Nvidia Grid, it seems that the GV100 is not on the required hardware list, however given its expensive nature, I wanted to double check prior to taking the plunge. So here is my question, can GV100 be passed to a Windows VM without any additional special license (I am not looking to share the GPU across other VMs).

TIA!

Hi

The GV100 is the Quadro equivalent of the V100 (Tesla). It supports Passthrough only or you can use it in a typical workstation tower chassis. It does not support vGPU and does not require any licensing.

If you wanted a Quadro GPU that also supports vGPU you’d need to use the Actively cooled version of the RTX6000 / RTX8000 or the A6000 (all 3 of those GPUs are Quadro).

Regards

MG

Thanks a bunch, that’s what I thought.