During its GPU Technology Conference today, Nvidia (TPG) to develop a CUDA C compiler for x86 platforms, which will be demonstrated in November at the SC10 Supercomputing conference.