Programming the LU factorization for a multicore system with accelerators

Jakub Kurzak; Piotr Luszczek; Mathieu Faverge; Jack Dongarra

doi:10.1007/978-3-642-38718-0_6

Programming the LU factorization for a multicore system with accelerators

Jakub Kurzak, Piotr Luszczek, Mathieu Faverge, Jack Dongarra

Computer Science and Mathematics Div

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Scopus citations

Abstract

LU factorization with partial pivoting is a canonical numerical procedure and the main component of the High Performance LINPACK benchmark. This article presents an implementation of the algorithm for a hybrid, shared memory, system with standard CPU cores and GPU accelerators. Performance in excess of one TeraFLOPS is achieved using four AMD Magny Cours CPUs and four NVIDIA Fermi GPUs.

Original language	English
Title of host publication	High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers
Pages	28-35
Number of pages	8
DOIs	https://doi.org/10.1007/978-3-642-38718-0_6
State	Published - 2013
Event	10th International Conference on High Performance Computing for Computational Science, VECPAR 2012 - Kobe, Japan Duration: Jul 17 2012 → Jul 20 2012

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7851 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	10th International Conference on High Performance Computing for Computational Science, VECPAR 2012
Country/Territory	Japan
City	Kobe
Period	07/17/12 → 07/20/12

Access to Document

10.1007/978-3-642-38718-0_6

Cite this

Kurzak, J., Luszczek, P., Faverge, M., & Dongarra, J. (2013). Programming the LU factorization for a multicore system with accelerators. In High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers (pp. 28-35). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7851 LNCS). https://doi.org/10.1007/978-3-642-38718-0_6

Kurzak, Jakub ; Luszczek, Piotr ; Faverge, Mathieu et al. / Programming the LU factorization for a multicore system with accelerators. High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers. 2013. pp. 28-35 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{aec7ac6377e045a9a95dbbc1d613c893,

title = "Programming the LU factorization for a multicore system with accelerators",

abstract = "LU factorization with partial pivoting is a canonical numerical procedure and the main component of the High Performance LINPACK benchmark. This article presents an implementation of the algorithm for a hybrid, shared memory, system with standard CPU cores and GPU accelerators. Performance in excess of one TeraFLOPS is achieved using four AMD Magny Cours CPUs and four NVIDIA Fermi GPUs.",

author = "Jakub Kurzak and Piotr Luszczek and Mathieu Faverge and Jack Dongarra",

year = "2013",

doi = "10.1007/978-3-642-38718-0_6",

language = "English",

isbn = "9783642387173",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "28--35",

booktitle = "High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers",

note = "10th International Conference on High Performance Computing for Computational Science, VECPAR 2012 ; Conference date: 17-07-2012 Through 20-07-2012",

}

Kurzak, J, Luszczek, P, Faverge, M & Dongarra, J 2013, Programming the LU factorization for a multicore system with accelerators. in High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7851 LNCS, pp. 28-35, 10th International Conference on High Performance Computing for Computational Science, VECPAR 2012, Kobe, Japan, 07/17/12. https://doi.org/10.1007/978-3-642-38718-0_6

Programming the LU factorization for a multicore system with accelerators. / Kurzak, Jakub; Luszczek, Piotr; Faverge, Mathieu et al.
High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers. 2013. p. 28-35 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7851 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Programming the LU factorization for a multicore system with accelerators

AU - Kurzak, Jakub

AU - Luszczek, Piotr

AU - Faverge, Mathieu

AU - Dongarra, Jack

PY - 2013

Y1 - 2013

N2 - LU factorization with partial pivoting is a canonical numerical procedure and the main component of the High Performance LINPACK benchmark. This article presents an implementation of the algorithm for a hybrid, shared memory, system with standard CPU cores and GPU accelerators. Performance in excess of one TeraFLOPS is achieved using four AMD Magny Cours CPUs and four NVIDIA Fermi GPUs.

AB - LU factorization with partial pivoting is a canonical numerical procedure and the main component of the High Performance LINPACK benchmark. This article presents an implementation of the algorithm for a hybrid, shared memory, system with standard CPU cores and GPU accelerators. Performance in excess of one TeraFLOPS is achieved using four AMD Magny Cours CPUs and four NVIDIA Fermi GPUs.

UR - http://www.scopus.com/inward/record.url?scp=84883285586&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-38718-0_6

DO - 10.1007/978-3-642-38718-0_6

M3 - Conference contribution

AN - SCOPUS:84883285586

SN - 9783642387173

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 28

EP - 35

BT - High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers

T2 - 10th International Conference on High Performance Computing for Computational Science, VECPAR 2012

Y2 - 17 July 2012 through 20 July 2012

ER -

Kurzak J, Luszczek P, Faverge M, Dongarra J. Programming the LU factorization for a multicore system with accelerators. In High Performance Computing for Computational Science, VECPAR 2012 - 10th International Conference, Revised Selected Papers. 2013. p. 28-35. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-38718-0_6

Programming the LU factorization for a multicore system with accelerators

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this