Mathematical models for evaluating efficiency and quality of means for synchronization of interacting processes in reconfigurable computer systems

• Alexey Martyshkin Penza State Technological University

Palabras clave: reconfigurable system, operating system, shared resource, process synchronization, task manager, model, stochastic network

Resumen

In this paper, the modelling method is applied to evaluate the performance characteristics of various methods for synchronization of parallel processes in order to determine which way to use to implement them: software or hardware. Comparative characteristics of process synchronization by the spin lock method and the core lock method are presented. Charts have been constructed that display the obtained characteristics.

Citas

Aliev T.I. (2009). Fundamentals of modelling discrete systems. St. Petersburg: St. Petersburg State University ITMO, 2009.363 p.

Andrews G.R. (2003). Basics of multi-threaded parallel and distributed programming. Transl. from English M. : Williams, 2003.512 p.

Biktashev R.A., Martyshkin A.I. (2013). A set of programs for measuring the performance of operating system functions.// XXI century: results of the past and problems of the present plus. 2013. No. 10 (14). Pp. 190-197.

Fundamentals of computational system theory (1978). Edited by S.A. Mayorov. M.: Higher School, 1978.408p.

Kleinrock L. (1979a). Computing systems with queues: Publishing House Mir - M. : 1979. 600 p.

Kleinrock L. (1979b). Queuing Theory. - M.: Mechanical Engineering, 1979. 432 p.

Kreutzler S. (1986). Designing operating systems for small computers: Translated from English.- M.: Mir, 1986. 680 p

Martyshkin A.I. (2016a). Mathematical modelling and performance evaluation of process synchronization tools in multiprocessor systems. Scientific notes of the International Humanitarian University. Papers of participants of the Second International Multidisciplinary Conference. 2016. P. 151-157.

Martyshkin A.I. (2016b). Semaphore performance issues in parallel systems. The evolution of modern science: Collection of papers of the International scientific-practical conference: in 3 parts. Executive Editor: Asatur Albertovich Sukiasyan.2016. P. 81-83.

Martyshkin A.I. (2016c). Modern methods for measuring the performance of multicore computing systems. New information technologies and systems; collection of scientific papers at the XIII International scientific and technical conference.2016. P. 128-131.

Martyshkin A.I. (2017). Performance evaluation of high-performance systems, taking into account failures and recoveries of computing nodes // Modern innovative technologies for training engineering personnel for mining and transport.2017. Number 4. P. 425-429.

Martyshkin A.I. (2018a). Possible structural organization and performance assessment of reconfigurable computing systems based on a common bus // Models, systems, networks in economics, technology, nature and society.2018. No. 1 (25). Pp. 141-150.

Martyshkin A.I. (2018b). Mathematical models of semaphores for coordinating access to shared resources of multiprocessor systems // Colloquium-journal.2018. No. 8-1 (19). Pp. 36-39.

Martyshkin, A.I. (2016d). Development and research of open-loop models the subsystem "processor-memory" of multiprocessor systems architectures UMA, NUMA and SUMA. ARPN Journal of Engineering and Applied Sciences. 2016. Volume 11. Issue 23. PP. 13526-13535

Martyshkin, A.I. (2016e). Mathematical modelling of Tasks Managers with the strategy in space with a homogeneous and heterogeneous input flow and finite queue. ARPN Journal of Engineering and Applied Sciences. 2016. Vol. 11. Issue 19 PP. 11325-11332.

Martyshkin, A.I., Yasarevskaya, O.N. (2015). Mathematical modelling of Task Managers for Multiprocessor systems on the basis of open-loop queuing networks. ARPN Journal of Engineering and Applied Sciences. 2015. Vol. 10. Issue 16 PP. 6744-6749.

Matalytsky M.A., Tikhonenko O.M., Koluzaeva E.V. (2011). Queuing systems and networks: analysis and requests: Monograph. Grodno: GrSU, 2011.816 p.

Tanenbaum E., Bos H. (2015). Modern Operating Systems. 4th ed. SPb. : Peter, 2015.1120 p.: ill. (Series "Classic computer science").