I am about to begin research with a university group using CUDA and GCGPU. We do not yet have the necessary NVIDIDA hardware, and I was wondering if there was a way to run CUDA and the software we create through some sort of emulated GPU. I tried to install CUDA on my machine at home (no NVIDIA card) and it said I had none of the necessary drivers. Is there a way to get around this? We will soon have a card (or hopefully a Tesla server through a grant), but we wanted to get a head start.