| Abstract [eng] |
The project is called \"Serial and parallel algorithms analysis in CUDA and OpenMP technologies“. The attention is paid to the aim of finding out OpenMP and CUDA technologies differences, capabilities and efficient configurations for most used algorithms. In fact, the study examines algorithms which can be operated in parallel and sequential order using OpenMP and CUDA technologies. They are tested with various configuration, data samples. It is expected that after the study of algorithms, their speed and other problems, connected with OpenMP and CUDA technologies, will be easier to determine how to further improve the previous half of the year generated code conversion system for these technologies. This project is also important because of computing resource utilization efficiency, the benefits of parallelism usage. Examination of sequential and parallel algorithms, visualized acceleration graphs will help to attract more emphasis on parallel programming by industry and educational institutions. |
