|
Article information
2014 , Volume 19, ¹ 4, p.84-98
Rakitskiy A.A., Ryabko B.Y., Fionov A.N.
The analytical method for comparing and evaluating the performance of computers and computer systems
In this report we discuss the results of application of a new method for evaluation of the performance of processing units and supercomputers. This method is based on notion of Computer Capacity, which is determined only by the computer architecture and can be computed theoretically on a design stage. Computer Capacity is based on the number of different tasks, which can be executed by computer in particular time. We calculated the computer capacity for different Intel and AMD processors list and supercomputers, based on these processors, and compared the results with the recognized benchmarks. This results show that Computer Capacity is fully objective characteristic and can be used for performance evaluation of processing units, especially on the design stage. Moreover, presented method can be used as alternative to traditional theoretical peak performance of supercomputers and could help developers of the computing systems. Aknowlegements: This work was supported by RFBR grants 12-07-00125-à. Received 27 November 2013.
[full text]
Keywords: Computer capacity, performance evaluation, supercomputers, benchmarks, computer systems
Author(s):
Rakitskiy Anton Andreevich
PhD. , Associate Professor
Office: Siberian State University of Telecommunications and Informatics
Address: 630102, Russia, Novosibirsk
E-mail: rakitsky.anton@gmail.com
SPIN-code: 7056-8681Ryabko Boris Yakovlevich
Dr. , Professor
Position: Head of Laboratory
Office: Federal Research Center for Information and Computational Technologies, Novosibirsk State University
Address: 630090, Russia, Novosibirsk, Academician M.A. Lavrentiev avenue, 6
Phone Office: (383) 334-91-24
E-mail: boris@ryabko.net
SPIN-code: 5580-5794Fionov Andrey Nikolaevich
Dr. , Professor
Position: Leading research officer
Office: Siberian St. Univ. of Telecom. and Information Sciences, Institute of Computational Technologies SB RAS
Address: 630102, Russia, Novosibirsk, Kirov St. 86
Phone Office: (383) 269 82 72
E-mail: fionov@sibsutis.ru
References:
[1] Ryabko B. An information-theoretic approach to estimate the capacity of processing units. Performance Evaluation. 2012; 69: 267-273.
[2] Shannon C.E. A mathematical theory of communication. Bell Sys. Tech. J. 1948; 27.
[3] Rakitskiy A.A., Ryabko B.Ja., Fionov A.N. Programmnyj kompleks dlya teoreticheskoj ocenki i analiza proizvoditel’nosti vychislitel’nyh system. Materialy konf. "Industrial’nye informacionnye sistemy". Novosibirsk. 2013. (In Russ.)
[4] Rakitskiy A.A. Prakticheskoe primenenie metodov teoreticheskoj ocenki vychislitel’noj sposobnosti dlja processorov Intel P5. Vestnik SibGUTI. 2012; 4: 50-61. (In Russ.)
[5] Rakitskiy A.A. Teoreticheskaja ocenka vychislitel’noj sposobnosti processorov Intel. Vestnik SibGUTI. 2013; 3: 29-45. (In Russ.)
[6] Programma dlja opredelenija vychislitel’noj sposobnosti i uravnenija vseh issledovannyh processorov. URL: https://drive.google.com/folderview?id=0B023P6Ajk7nJckhp RmdGS1VtZlk.
[7] TOP500. URL: http://top500.org.
[8] HPC Challenge Benchmark. URL: http://icl.cs.utk.edu/hpcc/.
[9] Intel x86 Quick Reference Instruction Manual - 8086/80186/80286/80386/80486. URL: http://www.intel-assembler.it/portale/5/x86-instruction-reference-manual/x86-instru-ction-reference-manual.asp.
[10] Fog A. The Microarchitecture of Intel, AMD and VIA CPUs an Optimization Guide for Assembly Programmers and Compiler Makers. Copenhagen Univ. College of Eng. 2012. URL: http://www.agner.org/optimize/.
[11] Fog A. Lists of instruction latencies, throughputs and microoperation breakdowns for Intel, AMD and VIA CPUs. Copenhagen Univ. College of Eng. 2012. URL: http://www.agner.org/optimize/.
[12] Intel 64 and IA-32 Architectures Software Developers Manual. Vol. 1. Basic Architecture. Intel Corp. URL: http://www.intel.ru/content/www/ru/ru/architecture-and-technology/64-ia-32-architectures-software-developer-vol-1-manual.html.
[13] Intel 64 and IA-32 Architectures Software Developers Manual. Vol. 2. Intel Corp. URL: http://www.intel.ru/content/www/us/en/architecture-and-technology/64-ia-32-architectures-software-developer-vol-2a-2b-instruction-set-a-z-manual.html.
Bibliography link:
Rakitskiy A.A., Ryabko B.Y., Fionov A.N. The analytical method for comparing and evaluating the performance of computers and computer systems // Computational technologies. 2014. V. 19. ¹ 4. P. 84-98
|
