I installed a GTX 570 in my Mac Pro. So far everything works great and it is recognized just fine. I left my GT 120 in there because I have the space and it is the native card. I know with Resolve/CUDA the recommended set up is run the displays off the GT 120 and use the faster card just for CUDA. Does OpenCL in Final Cut X work the same way? Honestly, I haven't noticed much of a difference in Final Cut (although Lux Mark and other benchmark programs shows a very large performance increase). I am currently still running the displays off the GT120 and wondering if that is the problem.