Exploiting Hyper-Loop Parallelism in Vectorization to Improve Memory Performance on CUDA GPGPU.

Shixiong Xu, David Gregg

Browse the full TrustCom paper archive.