Example Sobel Filter Thread Usage


Forgive me if my question is too easy or general to ask here:

I have begun to examine a code that is basically a different version of the Sobel Filter code given as an example CUDA SDK by the NVidia itself. ( http://www.nvidia.com/content/cudazone/cuda_sdk/samples.html ) However, I am really new in CUDA stuff, and I need to figure out how many threads are being used in that code. Can someone point out how sould I figure that out?

Another question, I have read the CUDA Manual by the NVidia (not all of it, but the initial parts) and do you guys know more resource/document about CUDA?