EVALUATION OF COMPUTATIONAL PRODUCTIVITY SCALING FOR PARALLEL IMPLEMENTATION OF LARGE EDDY SIMULAION METHOD
DOI:
https://doi.org/10.18372/2306-1472.50.127Keywords:
Large Eddy Simulation (LES), Mathematical modeling, Parallel multithreaded computations, turbulent flowsAbstract
For one of calculation-intensive methods of fluid-dynamics problems, which intensively becomes populardue to increasing power of high-performance computing and reducing the cost of computations, namely theLarge Eddy Simulation (LES) method, the structuring on elements and analysis of their ability to multithreadedparallelization have been performed. The advantages of explicit finite difference schemes incomparison with the implicit on the base of the criterion of efficiency in parallel algorithms andprogramming methods LES have been justified. A comparative analysis demonstrated that the mostdemanding of computer resources utilization is the solution of Poisson differential equations and the firstphase of the method of splitting by physical processes. The obtained results can be effectively used for bothjustification of the choice of calculation methodology and optimization of parallel algorithms for solving awide range of resource-demanding problems in different spheres of fluid-dynamics.References
Chapmann, D. 1979. Computational
aerodynamics, development and outlook. –
AIAA J. Vol. 17: P. 1293–1313.
Moin, P.; Kim, J. 1997. Tackling
Turbulence with Supercomputers. – Scientific
American. №276: P. 62–68.
Роуч П. Вычислительная гидромеханика /
П. Роуч / пер. с англ.. – М: Мир, 1980. – 616 с.
Ferziger, Joel H. 1996. Recent advances in
large eddy simulation. – Department of
mechanical engineering Stanford University.
Stanford CA: P. 1–13.
Шквар Є.О. Інтегрована гібридна техно-
логія паралельних обчислень / Є.О. Шквар //
Інтегровані технології та енергозбереження:
щокварт. наук.-практ. журн. НТУ (ХПІ). –
Харків: НТУ (ХПІ). – 2010. – № 1. –
С. 86–99.
Шквар Є.О. Гібридний метод паралель-
них обчислень / Є.О. Шквар // Вісник Чер-
каського університету: Сер. Прикладна ма-
тематика. Інформатика. – 2010. – Вип. 172. –
С. 123–136.