Parallelism through Graphics Processor Unit using MATLAB : A Survey

Divya Kundra

Abstract


While a CPU has handful number of cores, Graphics Processor Unit (GPU) has a large number of cores along with dedicated high speed memory. GPU allows hundreds of threads to run parallel on different cores thus accelerating algorithm by many folds. Many algorithms are speed up by GPU nowadays. Programs can be accelerated by GPU while using MATLAB more easily than by learning low level programming languages like C or Forton. This paper presents a survey study for using GPU through MATLAB. It discusses the high level programming language MATLAB, concept of parallelism, Parallel Computing Toolbox, GPU and the parallelism through GPU with help of MATLAB.

 

Keywords: Parallelism; Parallel Computing Toolbox, GPU, CPU, MATLAB Distributed Computing Server


Full Text:

PDF


DOI: https://doi.org/10.26483/ijarcs.v6i8.2591

Refbacks

  • There are currently no refbacks.




Copyright (c) 2016 International Journal of Advanced Research in Computer Science