Let's talk serious: Dynamic Programming - CUDA pt3

I've finally managed to get some time for learning CUDA. When it comes to a constant tutorial it's hard to find. I'm in a middle of Addisons Wesley's "CUDA by example". I must say it's a really good book, recommended by people that take their chances with CUDA. As it comes CUDA is a simple addition to C language with several rules. Everyrhing comes to a blocks of threads architecture. We can have X by Y blocks and X' by Y' threads. We can simply synchronize each blocks threads by shared arrays and __syncthreads() call. For now I don't have the knowledge and time to describe it more widely but I'm looking very forward to it. I'll try to do it in a form that anybody can review the document. I will concentrate on describing simple dynamic programming problem soled with CUDA. For now I'm going back to my master degree project. I'm trying to create a CUDA dll with a C call usage. Till next time!