I have dual 5970 video cards setup independantly (not crossfired) and two monitors connected one per card.
My idea is to boost my app performance by creating a GPU specific GL context, so I created my app that creates two display windows one on each monitor, in turn creating OpenGL context per GPU/card, in a way the context is GPU associated on which the display window was created, this I believe I achieved the GPU assosication (WGL_AMD_gpu_association) without needing to code anything in.
AMD claims this is equivalent of creating OGL context using API per this http://developer.amd.com/gpu/wgsdk/Documents/GPU_Association_WhitePaper.pdf
Can anyone here confirm its really the same when it comes to app performance?
Thanks in advance