Key concepts for parallel out-of-core LU factorization

Jack J. Dongarra; Sven Hammarling; David W. Walker

doi:10.1016/s0167-8191(96)00096-8

Key concepts for parallel out-of-core LU factorization

Jack J. Dongarra, Sven Hammarling, David W. Walker

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

16 Scopus citations

Abstract

This paper considers key ideas in the design of out-of-core dense LU factorization routines. A left-looking variant of the LU factorization algorithm is shown to require less I/O to disk than the right-looking variant, and is used to develop a parallel, out-of-core implementation. This implementation makes use of a small library of parallel I/O routines, together with ScaLAPACK and PBLAS routines. Results for runs on an Intel Paragon are presented and interpreted using a simple performance model.

Original language	English
Pages (from-to)	49-70
Number of pages	22
Journal	Parallel Computing
Volume	23
Issue number	1-2
DOIs	https://doi.org/10.1016/s0167-8191(96)00096-8
State	Published - Apr 1997
Externally published	Yes

Keywords

LU factorization
Out-of-core computation
Parallel I/O
Parallel computing

Access to Document

10.1016/s0167-8191(96)00096-8

Cite this

@article{e1f5dd81c8504b7fbe215857fbed8da6,

title = "Key concepts for parallel out-of-core LU factorization",

abstract = "This paper considers key ideas in the design of out-of-core dense LU factorization routines. A left-looking variant of the LU factorization algorithm is shown to require less I/O to disk than the right-looking variant, and is used to develop a parallel, out-of-core implementation. This implementation makes use of a small library of parallel I/O routines, together with ScaLAPACK and PBLAS routines. Results for runs on an Intel Paragon are presented and interpreted using a simple performance model.",

keywords = "LU factorization, Out-of-core computation, Parallel I/O, Parallel computing",

author = "Dongarra, \{Jack J.\} and Sven Hammarling and Walker, \{David W.\}",

year = "1997",

month = apr,

doi = "10.1016/s0167-8191(96)00096-8",

language = "English",

volume = "23",

pages = "49--70",

journal = "Parallel Computing",

issn = "0167-8191",

publisher = "Elsevier",

number = "1-2",

}

TY - JOUR

T1 - Key concepts for parallel out-of-core LU factorization

AU - Dongarra, Jack J.

AU - Hammarling, Sven

AU - Walker, David W.

PY - 1997/4

Y1 - 1997/4

N2 - This paper considers key ideas in the design of out-of-core dense LU factorization routines. A left-looking variant of the LU factorization algorithm is shown to require less I/O to disk than the right-looking variant, and is used to develop a parallel, out-of-core implementation. This implementation makes use of a small library of parallel I/O routines, together with ScaLAPACK and PBLAS routines. Results for runs on an Intel Paragon are presented and interpreted using a simple performance model.

AB - This paper considers key ideas in the design of out-of-core dense LU factorization routines. A left-looking variant of the LU factorization algorithm is shown to require less I/O to disk than the right-looking variant, and is used to develop a parallel, out-of-core implementation. This implementation makes use of a small library of parallel I/O routines, together with ScaLAPACK and PBLAS routines. Results for runs on an Intel Paragon are presented and interpreted using a simple performance model.

KW - LU factorization

KW - Out-of-core computation

KW - Parallel I/O

KW - Parallel computing

UR - https://www.scopus.com/pages/publications/0031123401

U2 - 10.1016/s0167-8191(96)00096-8

DO - 10.1016/s0167-8191(96)00096-8

M3 - Article

AN - SCOPUS:0031123401

SN - 0167-8191

VL - 23

SP - 49

EP - 70

JO - Parallel Computing

JF - Parallel Computing

IS - 1-2

ER -

Key concepts for parallel out-of-core LU factorization

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this