ATI Stream Processing vs NVIDIA CUDA what is the difference between the architectures.

Hi, I was wondering the real differences between the two architectures. Is there any paper of document which shows the differences between the architectures by using some kind of contraints like time, speed etc which metrically describes the differences.

Thanks!

U.K

CUDA and Stream are mere marketing terms, not architectures. Take a look at the OpenCL programming and best practices guides of NVIDIA and AMD; they describe the respective GPU architectures.