This is an implementation of CUDA eSTREAM-like framework. The code is copied from my old Master's thesis svn repo (so certain things might be missing -- if you find that this is the case, please let me know) at the request of a few students working on a similar topic. See the thesis for high-level details of the implementations.
The implementation includes an CUDA implementations of Grain, HC128, MICKEY, Rabbit, Salsa20, Trivium. Additionally an optimized (faster than the eSTREAM version) C implementation of Grain is provided.