Analysis of Parallel Computing Environment Overhead of OpenMP for Loop with Multi-core Processors

This paper analyze the parallel computing environment overhead of OpenMP for loop with multi-core processors including the case of data-race. The different solutions of data-race are discussed in present paper, such as critical pragma, atomic pragma and reduction clause. A new method is also presented in this paper which produces the least parallel environment overhead compared with other methods and the codes are presented too. The conclusions and discussions are beneficial to save computing time and improve the computational efficiency.

[1]  Barbara M. Chapman,et al.  John von Neumann Institute for Computing Enhancing OpenMP and Its Implementation for Programming Multicore Systems , 2008 .

[2]  Dhabaleswar K. Panda,et al.  Improving Scalability of OpenMP Applications on Multi-core Systems Using Large Page Support , 2007, 2007 IEEE International Parallel and Distributed Processing Symposium.

[3]  Santosh G. Abraham,et al.  Chip multithreading: opportunities and challenges , 2005, 11th International Symposium on High-Performance Computer Architecture.

[4]  Dirk Schmidl,et al.  Data and thread affinity in openmp programs , 2008, MAW '08.

[5]  Kunle Olukotun,et al.  A Single-Chip Multiprocessor , 1997, Computer.

[6]  Jung-Hsien Chiang,et al.  Neural and Fuzzy Methods in Handwriting Recognition , 1997, Computer.

[7]  Ami Marowka Performance of OpenMP Benchmarks on Multicore Processors , 2008, ICA3PP.

[8]  Dean M. Tullsen,et al.  Simultaneous multithreading: Maximizing on-chip parallelism , 1995, Proceedings 22nd Annual International Symposium on Computer Architecture.