I have two gpus: Tesla C870 and GeForce 8600GT, both in the same machine. I’m running a graphic application that uses OSG (Open Scenegraph) that is based on OpenGL. My intention is to perform graphic operations (execute the OSG instructions and OpenGL operations) using both gpus (or if it is not possible, perform the OSG and OpenGL operations using the Tesla) and show the graphic simulation into a monitor using the GeForce 8600GT, because Tesla doesn’t have video output. I have read in the Nvidia documentation, but I’m not sure about how to do it. Should I use the CUDA environment? Launching the OSG with some ‘nvidia-xconfig’ option, the graphics instructions will be performed in Tesla and visualized by Geforce 8600? Waiting for your replays. Thanks in advance.