UDC 004.272.26
INCREASED PERFORMANCE OF MULTIPROCESSOR COMPUTING SYSTEMS THROUGH BALANCED LOAD AND OPTIMUM MEMORY ACCESS
V. V. Filimonov, postgraduate student, RSREU, Ryazan, Russia;
orcid.org/0000-0001-6949-0579, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
V. V. Suskin, Dr. in technical sciences, full professor, RSREU, Ryazan, Russia;
orcid.org/0000-0001-5124-9672, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
Methods for evaluating performance and the possibility to use these methods for the development of parallel highly efficient algorithms running on multi-core processors and multi-processor computing systems are considered. The aim of this work is to study the existing methods for assessing performance to be used in choosing the optimal power of hardware for the implementation of computational tasks. The urgency of the problem lies in continuous growth of the amount of information that must be quickly processed and used to make key decisions in many technological areas. The paper considers the most important laws on which performance assessment is based, a numerical and graphical comparison of assessment methods is carried out, graphs of effective dependence of algorithms on hardware are obtained, and mathematical description of ensuring balanced load and optimal memory access is formulated.
Key words: performance of computing systems, optimization methods, multiprocessor computing systems, performance evaluation, efficiency of multiprocessor computing systems, performance improvement, parallelism, parallel computing.