Comparing Software Development Approaches for General Purpose GPU Computing

The Graphics Processing Unit (GPU) in today’s commodity video cards can be considered to be a powerful streaming co-processor on modern desktop workstations. This power enables tremendous increases in performance of certain applications by performing computations on the GPU instead of the CPU and also has a promise for new classes of interactive applications that take advantage of GPU computing power. However the GPU hardware and programming model is very restrictive which makes software development a complicated task. Several tools and techniques have been developed to alleviate this problem by abstracting hardware resources and generalizing the programming model. This paper examines approaches for general purpose software development on Programmable Graphics Processors and compares existing traditional and emerging software development systems.