ON PARALLELIZATION OF LINEAR SYSTEM EQUATION SOLVER IN FINITE ELEMENT SOFTWARE

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21110%2F17%3A00314270" target="_blank" >RIV/68407700:21110/17:00314270 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

ON PARALLELIZATION OF LINEAR SYSTEM EQUATION SOLVER IN FINITE ELEMENT SOFTWARE

Popis výsledku v původním jazyce

Numerical modelling using high performance computer brings in new opportunities in finite element computations. The contribution focuses on computing the efficiency of different existing approaches to solution of large, sparse, non-symmetric systems of linear equations, that can be found in any finite element software, the based on direct or iterative algorithms. In particular, direct solver from SuperLU library (X.S. Li, 1999), iterative solver from Portable, Extensible Toolkit for Scientific Computation (PETSc) are compared. SuperLU is a general-purpose library for the direct solution of large, sparse, non-symmetric systems of linear equations. The PETSc has successfully demonstrated that the use of modern programming paradigms can ease the development of large-scale scientific application codes (Satish Balay, 2016). The software has evolved into a powerful set of tools for the numerical solution of partial differential equations and related problems on high-performance computers. Additionally, the performance and scalability of parallel SuperLU solver is studied, using OpenMP (Chapman B, 2008, Barney B., 2014) and POSIX threads (B. Nicols, 1996) based on shared memory model. Also, the scalability of PETSc solver on distributed memory made using Message passing interface (MPI) (Peter S. Pacheco, 1997) is evaluated. The paper shows that parallelization can efficiently exploit the power of modern available hardware and significantly reducing the computation time.

Název v anglickém jazyce

ON PARALLELIZATION OF LINEAR SYSTEM EQUATION SOLVER IN FINITE ELEMENT SOFTWARE

Popis výsledku anglicky

Numerical modelling using high performance computer brings in new opportunities in finite element computations. The contribution focuses on computing the efficiency of different existing approaches to solution of large, sparse, non-symmetric systems of linear equations, that can be found in any finite element software, the based on direct or iterative algorithms. In particular, direct solver from SuperLU library (X.S. Li, 1999), iterative solver from Portable, Extensible Toolkit for Scientific Computation (PETSc) are compared. SuperLU is a general-purpose library for the direct solution of large, sparse, non-symmetric systems of linear equations. The PETSc has successfully demonstrated that the use of modern programming paradigms can ease the development of large-scale scientific application codes (Satish Balay, 2016). The software has evolved into a powerful set of tools for the numerical solution of partial differential equations and related problems on high-performance computers. Additionally, the performance and scalability of parallel SuperLU solver is studied, using OpenMP (Chapman B, 2008, Barney B., 2014) and POSIX threads (B. Nicols, 1996) based on shared memory model. Also, the scalability of PETSc solver on distributed memory made using Message passing interface (MPI) (Peter S. Pacheco, 1997) is evaluated. The paper shows that parallelization can efficiently exploit the power of modern available hardware and significantly reducing the computation time.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

20101 - Civil engineering

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2017

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of Engineering Mechanics 2017

ISBN

9788021454972

ISSN

—

e-ISSN

1805-8248

Počet stran výsledku

4

Strana od-do

198-201

Název nakladatele

Brno University of Technology

Místo vydání

Brno

Místo konání akce

Svratka

Datum konání akce

15. 5. 2017

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000411657600041