Hello,
I would like to ask some questions regarding corner cases when validating the GPU driver under virtualization environment:
- What kind of test would stress the GPU in ways that a non virtualized environment would not?
I mean from the perspective of the driver, not the hypervisor. I am not asking for specifics, just points to explore, I know this kind of knowledge it's not common and in most cases it's vendor specific, I just need some guidance on topics to study.
Thank you!