I am working on a dual card set up and I wanted to know if there is a way to tell which one to use? I am told that the equivalent in Cuda is Device Query.
Also what is the best way to get the computer to detect both cards with out having to plug in a monitor into both. I am using CentOS 5 as my operating system. We are curious what effects having a screen running on the card has on computation times.
Thanks for any help.
BRT_ADAPTER env variable. One process can use only 1 GPU core.
How do I change that on a linux system?