Accelerating JPEG Coding with Multiple Threads

Anyone interested in this needs to check out Fyodor Serzhenko's work: https://www.linkedin.com/company/fastvideo/

Doing the same thing but on CUDA cores. I am evaluating it for our needs and the quality of the whole SDK is outstanding. Great doco and testing, and performance is amazing. And thats not over 12 cores, it's over hundreds or thousands.