In the same vein, I'm wondering how much can be offloaded to the GPU for encoding, and with which software. I imagine Chris is the one with the knowledge on this. It's just curiosity on my part, though, since I don't intend to actually utilize the hardware for encoding at this time, but I do kind of want a new Nvidia card, if I can actually get ahold of one. CUDA is a compelling idea, but I don't know what the real-world implications are.