Performance comparison of various parallel incomplete LU factorization preconditioners for domain decomposition method

Sungwoo Kang; Hyounggwon Choi; Wanjin Chung; Yo Han Yoo; Jung Yul Yoo

doi:10.1007/s12206-018-1030-y

Performance comparison of various parallel incomplete LU factorization preconditioners for domain decomposition method

Sungwoo Kang, Hyounggwon Choi, Wanjin Chung, Yo Han Yoo, Jung Yul Yoo

Research output: Contribution to journal › Article › peer-review

4 Scopus citations

Abstract

A finite element code is parallelized by vertex-oriented domain decomposition method which utilizes one- or multi-dimensional partitioning in structured mesh and METIS Library in unstructured mesh. For obtaining the domain-decomposed solution, iterative solvers like conjugate gradient method are used. To accelerate the convergence of iterative solvers, parallel incomplete LU factorization preconditioners are employed, and their performances are compared. For the communication between processors, Message Passing Interface Library is used. The speedups of parallel preconditioned iterative solvers are estimated through computing 2- and 3-dimensional Laplace equations. The effects of mesh and partitioning method on the speedup of parallel preconditioners are also examined.

Original language	English
Pages (from-to)	5315-5323
Number of pages	9
Journal	Journal of Mechanical Science and Technology
Volume	32
Issue number	11
DOIs	https://doi.org/10.1007/s12206-018-1030-y
State	Published - 1 Nov 2018

Keywords

Domain decomposition method
Finite element method
Parallel ILU preconditioner
Preconditioned conjugate gradient

Access to Document

10.1007/s12206-018-1030-y

Cite this

@article{193a362597b24489b1c756e163e506d4,

title = "Performance comparison of various parallel incomplete LU factorization preconditioners for domain decomposition method",

abstract = "A finite element code is parallelized by vertex-oriented domain decomposition method which utilizes one- or multi-dimensional partitioning in structured mesh and METIS Library in unstructured mesh. For obtaining the domain-decomposed solution, iterative solvers like conjugate gradient method are used. To accelerate the convergence of iterative solvers, parallel incomplete LU factorization preconditioners are employed, and their performances are compared. For the communication between processors, Message Passing Interface Library is used. The speedups of parallel preconditioned iterative solvers are estimated through computing 2- and 3-dimensional Laplace equations. The effects of mesh and partitioning method on the speedup of parallel preconditioners are also examined.",

keywords = "Domain decomposition method, Finite element method, Parallel ILU preconditioner, Preconditioned conjugate gradient",

author = "Sungwoo Kang and Hyounggwon Choi and Wanjin Chung and Yoo, \{Yo Han\} and Yoo, \{Jung Yul\}",

note = "Publisher Copyright: {\textcopyright} 2018, The Korean Society of Mechanical Engineers and Springer-Verlag GmbH Germany, part of Springer Nature.",

year = "2018",

month = nov,

day = "1",

doi = "10.1007/s12206-018-1030-y",

language = "English",

volume = "32",

pages = "5315--5323",

journal = "Journal of Mechanical Science and Technology",

issn = "1738-494X",

number = "11",

}

TY - JOUR

T1 - Performance comparison of various parallel incomplete LU factorization preconditioners for domain decomposition method

AU - Kang, Sungwoo

AU - Choi, Hyounggwon

AU - Chung, Wanjin

AU - Yoo, Yo Han

AU - Yoo, Jung Yul

PY - 2018/11/1

Y1 - 2018/11/1

N2 - A finite element code is parallelized by vertex-oriented domain decomposition method which utilizes one- or multi-dimensional partitioning in structured mesh and METIS Library in unstructured mesh. For obtaining the domain-decomposed solution, iterative solvers like conjugate gradient method are used. To accelerate the convergence of iterative solvers, parallel incomplete LU factorization preconditioners are employed, and their performances are compared. For the communication between processors, Message Passing Interface Library is used. The speedups of parallel preconditioned iterative solvers are estimated through computing 2- and 3-dimensional Laplace equations. The effects of mesh and partitioning method on the speedup of parallel preconditioners are also examined.

AB - A finite element code is parallelized by vertex-oriented domain decomposition method which utilizes one- or multi-dimensional partitioning in structured mesh and METIS Library in unstructured mesh. For obtaining the domain-decomposed solution, iterative solvers like conjugate gradient method are used. To accelerate the convergence of iterative solvers, parallel incomplete LU factorization preconditioners are employed, and their performances are compared. For the communication between processors, Message Passing Interface Library is used. The speedups of parallel preconditioned iterative solvers are estimated through computing 2- and 3-dimensional Laplace equations. The effects of mesh and partitioning method on the speedup of parallel preconditioners are also examined.

KW - Domain decomposition method

KW - Finite element method

KW - Parallel ILU preconditioner

KW - Preconditioned conjugate gradient

UR - https://www.scopus.com/pages/publications/85057949559

U2 - 10.1007/s12206-018-1030-y

DO - 10.1007/s12206-018-1030-y

M3 - Article

AN - SCOPUS:85057949559

SN - 1738-494X

VL - 32

SP - 5315

EP - 5323

JO - Journal of Mechanical Science and Technology

JF - Journal of Mechanical Science and Technology

IS - 11

ER -

Performance comparison of various parallel incomplete LU factorization preconditioners for domain decomposition method

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this