As far as books go, I recommend two:
"CUDA by Example" and "Programming Massively Parallel Processors".
former is a straightforward intro while the other is a tad more technical.
The latter is also co-authored by the guy who runs the Coursera course.
I took the Udacity course last year, and I am taking the Coursera one now.
Though I am only half-way through the Coursera course, I would probably
recommend Udacity's. It is better structured, more intuitive and better
I think the most important is to get a feel for how CUDA works, and then it
really does not matter which language you use. Granted, the ways you check
for device properties and download/upload memory are slightly different,
but ultimately work the same way.
The mailing list, wiki, github gists and other online resources are there
to help you in case you need it.
Show replies by date