A paradigm shift in GP-GPU computing
Mehmet Esat Belviranli, Chih-Hsun Chou, Laxmi Narayan Bhuyan, Rajiv Gupta · 2014
General purpose computing on GPUs have became increasingly popular over the last decade. Scientific applications with SIMD computation characteristics show considerable performance improvements when run on these massively parallel architectures. However, data dependencies across thread blocks significantly impact the degree of achievable parallelism by requiring global synchronization across multi-processors (SMs) inside the GPU.