An approach for maximizing performance on heterogeneous clusters of CPU and GPU
Over the past years there has been significant enthusiasm for development of parallel computing on Graphics Processing Units (GPU) which have now become powerful and affordable hardware equipping data centers