Platformy pro paralelní zpracování úloh na GPU

Abstract

This thesis deals with the processing tasks to the graphics card. Specific types of tasks are selected sorting algorithms. The first part includes description CUDA and OpenCL technology in which sorting algorithm is implemented. Next it is described the principle of the algorithm and its implementation. Next step is profiling and optimization of sorting algorithm. The last part includes testing these algorithms on different graphics cards and a comparison of both technologies.