Continued effectiveness of the calculation of the total recognition for graphics processor

Main Article Content

I.P. Strubytsʹka

Abstract

he test to compare the time of execution of the same programs at the central and graphics processors, matrix multiplication in particular, is conducted in the article. CUDA technology is used to solve the problem. Productivity, speeding and efficiency of the parallel system is calculated

Article Details

How to Cite
Strubytsʹka, I. . (2011). Continued effectiveness of the calculation of the total recognition for graphics processor. Electronics and Communications, 16(3), 117–120. https://doi.org/10.20535/2312-1807.2011.16.3.266227
Section
Electronic systems

References

CUDA Zone — The resource for CUDA developer, http://www.nvidia.com/object/cuda_home_new.html

Using graphics cards for computing, http://GPGPU.ru

NVIDIA CUDA Compute Unified Device Architecture, Programming Guide, Version 2.0, 2008, p.107

A. Boreskov and A. Kharlamov, Basicswork with CUDA technology, M.: DMKPress, 2010, p. 232.

Y. Kozak, P. Stakhiv, and I. Strubitska, “Parallelization of the optimization algorithm parameters for discrete dynamic models on massively parallel processors”, Selection and processinginformation, vol. 32, no. 108, pp. 126–130, Jan. 2010.

L. Ljung, Identification of systems. Theory foruser, M.: Science, 1991, p. 432.

A. Semenov, D. Artamonov, and A. Bryukhachev, Identification of control objects: Textbook. by allowance, Penza: Penz Publishing House. state un-ta, 2003, p. 211.

L. Rastrigin, Adaptation of complex systems, Riga: Zinatne, 1981, p. 375.

P. Stakhiv and Y. Kozak, “Pobudovamacromodels of electromechanical componentswith vikoristannym optimization”, Tekhnіchnaelectrodynamics, no. 4, pp. 33–36, 2001.