The use of CPU and GPU for calculations in Matlab


The article presents selected solutions using CPU processors and GPUs for calculations in the Matlab environment. Various methods of performing calculations on the CPU as well as on the GPU were compared. Differences, disadvantages, advantages and effects of using selected calculation methods have been indicated.


CPU; GPU; Matlab

[1] MATLAB Product Description - MathWorks Documentation, [01.08.2018]
[2] K. Banasiak, Algorytmizacja i programowanie w MATLABIE, Wydawnictwo BTC, 2017.
[3] Parallel Computing Toolbox - Documentation, [01.08.2018].
[4] J. W. Sut, Y. Kim, MATLAB and Parallel Computing Toolbox, 2014, 99-125.
[5] Obliczenia równoległe w środowisku Matlab - MathWorks Video and Webinars, [01.08.2018].
[6] B. Mrozek, „Obliczenia równoległe w Matlab-ie,” Pomiary Automatyka Robotyka, tom R. 15, nr 2, pp. 285-294, 2011.
[7] I. Azzini, R. Muresano, M. Ratto, Dragonfly: A multi-platform parallel toolbox for MATLAB/Octave, 2018, 21-42.
[8] What is GPU computing?, [01.08.2018].
[9] M. Sourouri, J. Langguth, F. Spiga, S. B. Baden. X. Cai, CPU+GPU Programming of Stencil Computations for Resource-Efficient Use of GPU Clusters, 2015, 17-26.
[10] Using parfor-loop - MathWorks Documentation, [01.08.2018].
[11] What Is Parallel Computing?, [01.08.2018].
[12] Using FFT on the GPU for Spectral Analysis MathWorks –Documentation, [01.08.2018].
[13] H. Anzt, M. Gates, J. Dongarra, M. Kreutzer, G. Welling, M. Kohler, Preconditioned Krylov solvers on GPUs, 2017, 32-44.

Published : 2019-03-30

Woźniak, J. (2019). The use of CPU and GPU for calculations in Matlab . Journal of Computer Sciences Institute, 10, 32-35.

Jarosław Woźniak
Lublin University of Technology  Poland