Abstract
The paper considers the problem of developing rational methods for the creation of a framework (a plan, execution timetable) of parallel programs for real parallel computing systems. To solve this problem, a software environment (software stand) has been developed that allows implementing different strategies for building a framework for parallel programs and assessing the quality of these strategies. The built-in script Lua programming language is used to increase the flexibility of modeling and optimization capabilities. Results of applying some of the proposed strategies for constructing rational plans for parallel programming are outlined.
References
[1] Voevodin V. V., Voevodin Vl. V., Parallel’nye vychislenija (Parallel computing), St. Petersburg, BHV-Petersburg, 2002 (in Russian)Search in Google Scholar
[2] Ierusalimschy R., Programming in Lua, 3rd Edition, PUC-Rio, Brasil, Rio de Janeiro, 2013Search in Google Scholar
[3] AlgoWiki, Open Encyclopedia of Parallel Algorithmic Features, Available at: 18.10.2016, http://algowiki-project.org/en/Open_Encyclopedia_of_Parallel_Algorithmic_Features (reference date: 01.06.2018)Search in Google Scholar
[4] Gary M., Johnson D., Vychislitel’nye mashiny i trudnoreshaemye zadachi (Computers and intractable problems), Moscow, Mir, 1982 (in Russian)Search in Google Scholar
[5] Fisher J. A., Very long instruction word architectures and the ELI- 512, In: Proceedings of the 10th annual International Symposium on Computer Architecture, New York, NY, USA: Association for Computing Machinery (ACM), 1983, 140-15010.1145/1067651.801649Search in Google Scholar
[6] McNairy C., Soltis D., Itanium2 processor microarchitecture, IEEE Micro Magazine, 2003, 23(2), 44-5510.1109/MM.2003.1196114Search in Google Scholar
[7] Bakanov V. M., Dynamics control computing in the processor data flow architecture for different types of algorithms, Programmnaya Ingeneria (Software Engineering), 2015, 9, 20-24 (in Russian).Search in Google Scholar
© by Valery Bakanov, published by De Gruyter
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.