Hi,

Not exactly my area and I’m lost in a sea of solutions, I need help.
There are so many out there but I don’t know any of them or if they are still maintained, offer full solution, time to instance up, etc.

Problem is simple to describe.
I want to setup access to GPU instances in order to run any python code the project devs have built.
The hardware consists of several servers with GPUs that support vGPU, NVIDIA GPU virtualization solution.

I’m looking for something similar to https://www.runpod.io/

What open source software can be used to spawn the client machines from the existing hardware pool?

I’m looking into kubernetes for automation and MAAS from canonical for the rest. Am I missing something important?

Any help or insight would be helpful.