METHODS FOR EVALUATING THE PERFORMANCE OF HETEROGENEOUS COMPUTER SYSTEMS
https://doi.org/10.32362/2500-316X-2017-5-6-11-19
Abstract
Article is devoted to the performance evaluation of heterogeneous computing systems (platforms). The article substantiates the relevance of the use of problem-oriented configuration of computational systems for automated systems of information processing and management. The proposed mechanism is parallel and pipeline processing allowing to parallelize the task and thereby improve system performance. For interaction between modules it was proposed to use MPI-message passing interface between processes that perform a single task. The performance measurement was performed using recognized test kits Intel “MPI Benchmark” and “LINPACK HPL”, which allow to fully assess the effectiveness of the computing system and its relevance for the user. Given the graphics performance of the heterogeneous computing complex “profit center” in dealing with complex problems of linear algebra. The results obtained confirm high efficiency of MVP for various applications.
Keywords
вычислительная система,
тест производительности,
оценка производительности,
управление параллелизмом операций,
характеристики вычислительных систем,
вычислительная система,
тест производительности,
оценка производительности,
управление параллелизмом операций,
характеристики вычислительных систем
About the Authors
S. A. Sorokin
M.A. Kartsev Scientific and Research Institute of Computing Systems
Russian Federation
Russian Federation
M. Z. Benenson
M.A. Kartsev Scientific and Research Institute of Computing Systems
Russian Federation
Russian Federation
A. P. Sorokin
M.A. Kartsev Scientific and Research Institute of Computing Systems
Russian Federation
Russian Federation
References
1. Sorokin S.A., Chudinov S.M., Sorokin A.P., Bolgova E.V. Methods for evaluating the performance of computer systems // Nauchnje vedomosti BelGU. Economika. Informatika. (Belgorod State University Scientic Bulletin). 2017. № 9(258). V. 42. P. 89–97. (in Russ.)
2. Bychkov I.N., Molchanov I.A., Feldman V.M., Yurlin S.V. Computing microprocessor complexes with “Elbrus” and SPARC architectures for the automated control systems creation // Kachestvo i zhizn' (Quality and Life). 2016. № 4(12). P. 74–81. (in Russ.).
3. Galagan P.V. The GRYPHON platform for solving problems of embedded systems special purpose // Sovremennje technologii avtomatizachii (Contemporary Technologies in Automation).2015. № 4(77). P. 16–23. (in Russ.).
4. Ваrybin A.K., Lobanov V.N., Cheldiev M.I., Chuchkalov P.V. A reconfiguration computing platform with diverse architecture // Voprosy radioelectroniki (Questions of radio electronics). 2016. № 7. P. 70–77. (in Russ.).
5. Sorokin A.P. Chudinov S.M. Scientific approach to the election of designing computing platforms for image processing // Radiopromeshlennost' (Radio Industry). 2017. № 3. P. 68– 73. (in Russ.).
6. Gropp W., Lusk E., Skjellum A. Using MPI: Portable parallel programming with the message passing interface (Scientific and Engineering Computation). 2nd Ed. MIT Press, 1999. 386 p.
7. Dongarra J.J. Performance of various computers using standard linear equations software. // Electrical Engineering and Computer Science Department University of Tennessee CS- 89-85 September 9, 2007.
8. Sanders J., Kandrot E. CUDA by Example: An Introduction to General-Purpose GPU Programming. M.: DMK Press Publ., 2011. 232 p.
Review
For citations:
Sorokin S.A., Benenson M.Z., Sorokin A.P. METHODS FOR EVALUATING THE PERFORMANCE OF HETEROGENEOUS COMPUTER SYSTEMS. Russian Technological Journal. 2017;5(6):11-19. (In Russ.) https://doi.org/10.32362/2500-316X-2017-5-6-11-19
Views:
549
Email this article 