UDC 004.424.5.032.24
PARALLEL ALGORITHMS OF DATA SORTING AND THEIR IMPLEMENTATION ON CUDA PLATFORM
S. V. Skvortsov, PhD (technical sciences), full professor, RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
T. A. Pyurova, undergraduate, RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
Theoretical and practical implementation of algorithms for parallel data sorting based on CUDA platform is considered. The relevance of this topic is due to the constant increase of the volume of sorted data at present time, and the demand to increase speed of information processing. The aim of this work is a modification of known sorting algorithms for their effective use on CUDA platform, i.e. it is necessary to develop a program of parallel data sorting to compare its performance with sequential version and systematize the results. Thus, obtained results show that the use of graphics accelerator can significantly increase the speed of applications that use the algorithms of data sorting. With the growth of COP-operated arrays and the size of used blocks of work time, multi-threaded application on GPU is reduced several tens of times compared to the sequential implementation on the CPU, and with the increase in the number of blocks of winning increases.
Key words: device, grid, kernel, graphic memory, block, thread, host, sort.