I’m currently working at a company that is planning to set up an NVIDIA OVX server. However, we have a couple of questions that we could not find any answer to them on the NVIDIA website.
Our main concern is the deployment process for the server, we are wondering if anyone in the community could provide us with a manual or guidelines for setting up an OVX server.
Additionally, we’re curious about the possibility of running multiple instances of Omniverse on this server given that it comes with 8 GPUs. Specifically, we’re wondering if it’s feasible for 8 team members to work on separate Omniverse instances simultaneously, with each person utilizing their dedicated GPU. Could anyone shed some light on how GPU memory is shared in such setups?
Lastly, we’re also interested in understanding more about parallelization when it comes to deploying Omniverse instances on multiple GPUs. How can we optimize resource usage and performance in this context?