Research of manual and automated optimization approaches for typical algorithmic structures for graphics processors |
| 1 | |
| 2013 |
| scientific article | 004.272.3 | ||
| 289-294 | graphical accelerators, parallell computing, optimizations. |
| The work is devoted to implementation of a set of typical computational algorithms on GPU and research on how to optimize them. The results show an efficiency of different optimizations applied to various kernels. Kernels were tested on NVIDIA and AMD GPUs. Mathematical kernels were tested in single and double precision mode. |
 |