low-level server clustering, and sharing gpu compute dynamically between virtual machines.

Reply to this note

Please Login to reply.

Discussion

so solutions to pay more but pay less in other terms

Not sure I follow. For vgpu it's resource sharing. So I can run my LLM servers alongside my GUI desktop VMs.

I thought you had to buy something new

Well technically you need recent very nvidia grid gpus for it work officially. But there are tools that let you run on older tesla cards (which I have) or even consumer nvidia cards.