Forgive, me this has probably been covered before but I can’t find any references.
I want to dedicate one card to running numerically intensive programs and another to graphics.
Please help me with instructions or direct me to a topic
What operating system are you running?
I have been doing this for quite a while. I run a GT 640 with my displays and have used a Tesla K20 and now a GTX Titan as my compute card.
The relevant question is what application are you running on the compute card? Something you compiled? Some app that someone else coded? If it’s the former and it’s a straight CUDA app, you can do something like this:
[url]https://www.cs.virginia.edu/~csadmin/wiki/index.php/CUDA_Support/Choosing_a_GPU[/url]
If it’s an app that someone else developed, you’ll have to look at the documentation to see if there is an option to make use of a specific GPU with said application.
I am running Centos 6.3