I'm seeking guidance on utilizing a VM in conjunction with Multi-Instance GPU (MIG) technology. I've come across an article detailing its application specifically with the A100 GPU and Vsphere 7, but I'm interested in exploring its compatibility with Vsphere 8 and the H100 GPU. My goal is to allocate CPU, memory, storage, and GPU resources to VMs. Could you kindly provide insights on the feasibility of this endeavour?
If it's viable, I'd greatly appreciate a step-by-step guide on implementation. I intend to utilize PCIe GPUs equipped with NvAI licenses. Are there any additional prerequisites or considerations beyond GPU and the hypervisor, or will they suffice to manage the setup seamlessly?
Thank you in advance for your assistance.
Best Regards
Arun Sharma
