I am trying to create an OpenGL program that will run on a cluster that has Kepler cards attached to each node. The compute nodes on the cluster does not have any monitor connected to it and there is no xorg.conf file though there is an nvidia driver so that cuda can be used.
My question is how to I create an OpenGL context in this case to run an OpenGL program? Is it possible to have an OpenGL program without a context, as far as I know, the answer is no!. (Is this assumption wrong???)
What do I need on the cluster so that I can get an OpenGL context? Can I create a dummy xorg.conf file? Will that work?