Evolution and trends in GPU computing

Central Processing Units (CPUs) are task-parallel, latency-oriented processors, while Graphics Processing Units (GPUs) are data-parallel, throughput oriented processors. Besides their traditional use as graphics coprocessors, the GPUs have been used in recent years for general purpose computations, too. Rapid development of graphics hardware led to an extensive use in both scientific and commercial applications. Numerous papers report high speedups in various domains. This paper presents an effort to bring GPU computing closer to programmers and wider community of users. GPU computing is explored through NVIDIA Compute Unified Device Architecture (CUDA) that is currently the most mature application programming interface (API) for general purpose computation on GPUs.