Hello world! Can vGPU uses GPUs from multiple machines?

Hello world! First post here.

In a two-node HCI structure, if each node is a bare metal with vGPU capable GPUs, can vGPU profiles use resources from both bare metals?


You cannot use vGPU with baremetal at all. In addition, it is not possible to assign GPU resources from different hosts to a single VM.



I was meant to say each bare metal’s GPU. Bare metal OS is EXSi.

The use case is a 2-node HA cluster for remote CAD work. Identical bare metal spec. Each machine has one GPU. Perhaps better to have two GPUs on the same machine? But what if that machine fails?