Hello,
I would like to now how is CUDA supported in Windows server environment (more specifically, WIndows HPC server 2008).
From what I can find, very new GPU adapters come with driver for Windows server, and from the system requirement of CUDA
Zone, Windows servers are not listed.
If Windows servers are supported, I have yet another question. How to make the HPC scheduler arrange GPU properly,
especially when there are more than 1 GPU available?