Performance Analysis and Optimisation of Two-sided Factorization Algorithms for Heterogeneous Platform

Many applications, ranging from big data analytics to nanostructure designs, require the solution of large dense singular value decomposition (SVD) or eigenvalue problems. A first step in the solution methodology for these problems is the reduction of the matrix at hand to condensed form by two-side...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Procedia computer science Jg. 51; S. 180 - 190
Hauptverfasser: Kabir, Khairul, Haidar, Azzam, Tomov, Stanimire, Dongarra, Jack
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 2015
Schlagworte:
ISSN:1877-0509, 1877-0509
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Many applications, ranging from big data analytics to nanostructure designs, require the solution of large dense singular value decomposition (SVD) or eigenvalue problems. A first step in the solution methodology for these problems is the reduction of the matrix at hand to condensed form by two-sided orthogonal transformations. This step is standardly used to significantly accelerate the solution process. We present a performance analysis of the main two-sided factorizations used in these reductions: the bidiagonalization, tridiagonalization, and the upper Hessenberg factorizations on heterogeneous systems of multicore CPUs and Xeon Phi coprocessors. We derive a performance model and use it to guide the analysis and to evaluate performance. We develop optimized implementations for these methods that get up to 80% of the optimal performance bounds. Finally, we describe the heterogeneous multicore and coprocessor development considerations and the techniques that enable us to achieve these high-performance results. The work here presents the first highly optimized implementation of these main factorizations for Xeon Phi coprocessors. Compared to the LAPACK versions optmized by Intel for Xeon Phi (in MKL), we achieve up to 50% speedup.
AbstractList Many applications, ranging from big data analytics to nanostructure designs, require the solution of large dense singular value decomposition (SVD) or eigenvalue problems. A first step in the solution methodology for these problems is the reduction of the matrix at hand to condensed form by two-sided orthogonal transformations. This step is standardly used to significantly accelerate the solution process. We present a performance analysis of the main two-sided factorizations used in these reductions: the bidiagonalization, tridiagonalization, and the upper Hessenberg factorizations on heterogeneous systems of multicore CPUs and Xeon Phi coprocessors. We derive a performance model and use it to guide the analysis and to evaluate performance. We develop optimized implementations for these methods that get up to 80% of the optimal performance bounds. Finally, we describe the heterogeneous multicore and coprocessor development considerations and the techniques that enable us to achieve these high-performance results. The work here presents the first highly optimized implementation of these main factorizations for Xeon Phi coprocessors. Compared to the LAPACK versions optmized by Intel for Xeon Phi (in MKL), we achieve up to 50% speedup.
Author Haidar, Azzam
Dongarra, Jack
Kabir, Khairul
Tomov, Stanimire
Author_xml – sequence: 1
  givenname: Khairul
  surname: Kabir
  fullname: Kabir, Khairul
  email: kkabir@eecs.utk.edu
  organization: University of Tennessee, Knoxville, TN, USA
– sequence: 2
  givenname: Azzam
  surname: Haidar
  fullname: Haidar, Azzam
  email: haidar@eecs.utk.edu
  organization: University of Tennessee, Knoxville, TN, USA
– sequence: 3
  givenname: Stanimire
  surname: Tomov
  fullname: Tomov, Stanimire
  email: tomov@eecs.utk.edu
  organization: University of Tennessee, Knoxville, TN, USA
– sequence: 4
  givenname: Jack
  surname: Dongarra
  fullname: Dongarra, Jack
  email: dongarra@eecs.utk.edu
  organization: University of Tennessee, Knoxville, TN, USA
BookMark eNqFkEFPwyAUx4mZiXPuE3jhC7QCXUt78LAszpks2Q67k1egk6UtC6Bmfnrp5sF40HfhvUd-_8DvFo1622uE7ilJKaHFwyE9Oit9ygjNU5KnjLErNKYl5wnJSTX60d-gqfcHEisry4ryMWq22jXWddBLjec9tCdvPIZe4c0xmM54CMb22DZ492ETb5RWeAkyWGc-L1fzdh-H8Np5HIPwSgft7F732r55vG0hDPF36LqB1uvp9zlBu-XTbrFK1pvnl8V8nchsVoYEWKb4DKpMZTmjUBVKcgK0zssCiCRNwXhcKwkNY4pVnNeS5kU9o7qgtZLZBFWXWOms9043QppwfmZwYFpBiRiUiYM4KxODMkFyEZVFNvvFHp3pwJ3-oR4vlI6_ejfaCS-NjjKVcVoGoaz5k_8CmWSLeQ
CitedBy_id crossref_primary_10_1007_s10586_015_0515_z
Cites_doi 10.1007/s10543-008-0180-1
10.1007/BF02161358
10.1137/0612026
10.1093/comjnl/4.4.332
10.1145/1055531.1055539
10.1016/S0167-8191(99)00021-6
10.1145/365723.365736
10.1137/0910005
10.1177/1094342013502097
10.1145/44128.44130
10.1137/0908009
10.1093/comjnl/3.1.23
10.1016/j.parco.2010.06.001
ContentType Journal Article
Copyright 2015 The Authors
Copyright_xml – notice: 2015 The Authors
DBID 6I.
AAFTH
AAYXX
CITATION
DOI 10.1016/j.procs.2015.05.222
DatabaseName ScienceDirect Open Access Titles
Elsevier:ScienceDirect:Open Access
CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 1877-0509
EndPage 190
ExternalDocumentID 10_1016_j_procs_2015_05_222
S1877050915010303
GroupedDBID --K
0R~
0SF
1B1
457
5VS
6I.
71M
AACTN
AAEDT
AAEDW
AAFTH
AAIKJ
AALRI
AAQFI
AAXUO
ABMAC
ACGFS
ADBBV
ADEZE
AEXQZ
AFTJW
AGHFR
AITUG
ALMA_UNASSIGNED_HOLDINGS
AMRAJ
E3Z
EBS
EJD
EP3
FDB
FNPLU
HZ~
IXB
KQ8
M41
M~E
NCXOZ
O-L
O9-
OK1
P2P
RIG
ROL
SES
SSZ
9DU
AAYWO
AAYXX
ABWVN
ACRPL
ACVFH
ADCNI
ADNMO
ADVLN
AEUPX
AFPUW
AIGII
AKBMS
AKRWK
AKYEP
CITATION
~HD
ID FETCH-LOGICAL-c348t-a23d74a93d3521a96dc70a1b586a0c0f62721adcaf22d2977bc156b41e61bdc3
ISICitedReferencesCount 1
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000373939100018&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1877-0509
IngestDate Tue Nov 18 22:24:16 EST 2025
Sat Nov 29 02:44:38 EST 2025
Wed May 17 01:33:33 EDT 2023
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Keywords Multicore
Xeon Phi
Task-based programming
Eigensolver
Language English
License http://creativecommons.org/licenses/by-nc-nd/4.0
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c348t-a23d74a93d3521a96dc70a1b586a0c0f62721adcaf22d2977bc156b41e61bdc3
OpenAccessLink https://dx.doi.org/10.1016/j.procs.2015.05.222
PageCount 11
ParticipantIDs crossref_citationtrail_10_1016_j_procs_2015_05_222
crossref_primary_10_1016_j_procs_2015_05_222
elsevier_sciencedirect_doi_10_1016_j_procs_2015_05_222
PublicationCentury 2000
PublicationDate 2015
2015-00-00
PublicationDateYYYYMMDD 2015-01-01
PublicationDate_xml – year: 2015
  text: 2015
PublicationDecade 2010
PublicationTitle Procedia computer science
PublicationYear 2015
Publisher Elsevier B.V
Publisher_xml – name: Elsevier B.V
References Christian, Bruno Lang, Xiaobai (bib0015) 2000; 26
Jack, Danny, Sven (bib0030) 1989; 27
Geist George Howell, Gary W. and T Rowan. Error analysis of reduction to banded hessenberg form.
Intel. Math kernel library. https://software.intel.com/en-us/en-us/intel-mkl/.
Bischof, van Loan (bib0010) 1987; 8
Roger, Horst (bib0045) September 1988; 14
John McCalpin. STREAM: Sustainable Memory Bandwidth in High Performance Computers. (http://www.cs.virginia.edu/stream/).
Khairul Kabir, Azzam Haidar (bib0070) 2014
Francis (bib0035) 1961; 4
Bo K˚agström, Daniel Kressner, Enrique Quintana-Orti, Gregorio Quintana-Orti (bib0075) 2008; 48
Larss Karlsson, Bo (bib0080) 2011
Martin, Wilkinson (bib0090) 1968; 12
Wilkinson (bib0120) 1960; 3
Dongarra, Gates, Haidar, Kabir, Luszczek, Tomov, Yamazaki (bib0020) 2014
Bruno Lang. (bib0085) 1999; 25
George Geist (bib0040) 1991; 12
Anderson, Bai, Bischof, Suzan, James, Jack, Du CrozF J., Greenbaum, Hammarling, Danny (bib0005) 1999
Stanimire Tomov, Rajib Nath, Jack Dongarra (bib0110) December 2010; 36
Heinz (bib0100) 1958; 49
Haidar, Tomov, Dongarra, Solca, Schulthess (bib0050) May 2014; 28
Jack, Cleve (bib0025) 1984
Wachspress (bib0115) 1995
Gary, Nadia (bib0055) March 2005; 31
Rep. ORNL/TM-13344.
Schreiber, van Loan (bib0105) 1991; 10
Bo K˚agström (10.1016/j.procs.2015.05.222_bib0075) 2008; 48
Jack (10.1016/j.procs.2015.05.222_bib0025) 1984
Stanimire Tomov (10.1016/j.procs.2015.05.222_bib0110) 2010; 36
10.1016/j.procs.2015.05.222_bib0095
Anderson (10.1016/j.procs.2015.05.222_bib0005) 1999
Francis (10.1016/j.procs.2015.05.222_bib0035) 1961; 4
10.1016/j.procs.2015.05.222_bib0065
Khairul Kabir (10.1016/j.procs.2015.05.222_bib0070) 2014
Christian (10.1016/j.procs.2015.05.222_bib0015) 2000; 26
Gary (10.1016/j.procs.2015.05.222_bib0055) 2005; 31
Bruno Lang. (10.1016/j.procs.2015.05.222_bib0085) 1999; 25
Jack (10.1016/j.procs.2015.05.222_bib0030) 1989; 27
10.1016/j.procs.2015.05.222_bib0060
Martin (10.1016/j.procs.2015.05.222_bib0090) 1968; 12
Heinz (10.1016/j.procs.2015.05.222_bib0100) 1958; 49
Wachspress (10.1016/j.procs.2015.05.222_bib0115) 1995
Haidar (10.1016/j.procs.2015.05.222_bib0050) 2014; 28
Roger (10.1016/j.procs.2015.05.222_bib0045) 1988; 14
Wilkinson (10.1016/j.procs.2015.05.222_bib0120) 1960; 3
George Geist (10.1016/j.procs.2015.05.222_bib0040) 1991; 12
Bischof (10.1016/j.procs.2015.05.222_bib0010) 1987; 8
Dongarra (10.1016/j.procs.2015.05.222_bib0020) 2014
Schreiber (10.1016/j.procs.2015.05.222_bib0105) 1991; 10
Larss Karlsson (10.1016/j.procs.2015.05.222_bib0080) 2011
References_xml – volume: 27
  start-page: 215
  year: 1989
  end-page: 227
  ident: bib0030
  article-title: Block reduction of matrices to condensed forms for eigenvalue computations
  publication-title: Journal of Computational and Applied Mathematics
– year: 2014
  ident: bib0070
  article-title: Stanimire Tomov, and Jack Dongarra. Performance analysis and design of a Hessenberg reduction using stabilized blocked elementary transformations on new architectures
  publication-title: Computer science dept. technical report, University of Tennessee, November
– volume: 4
  start-page: 332
  year: 1961
  end-page: 345
  ident: bib0035
  article-title: The QR transformation, part 2
  publication-title: Computer Journal
– volume: 3
  start-page: 23
  year: 1960
  end-page: 27
  ident: bib0120
  article-title: Householder's method for the solution of the algebraic eigenproblem
  publication-title: The Computer Journal
– year: 1999
  ident: bib0005
  article-title: LAPACK User's Guide. Society for Industrial and Applied Mathematics, Philadelphia
– start-page: 68
  year: 1984
  end-page: 87
  ident: bib0025
  publication-title: EISPACK: A package for solving matrix eigenvalue problems.
– reference: John McCalpin. STREAM: Sustainable Memory Bandwidth in High Performance Computers. (http://www.cs.virginia.edu/stream/).
– volume: 10
  start-page: 53
  year: 1991
  end-page: 57
  ident: bib0105
  article-title: A storage-efficient WY representation for products of Householder transformations
  publication-title: J. Sci. Stat. Comput.
– volume: 48
  start-page: 563
  year: 2008
  end-page: 584
  ident: bib0075
  article-title: Blocked Algorithms for the Reduction to Hessenberg-Triangular Form Revisited
  publication-title: BIT Numerical Mathematics
– volume: 36
  start-page: 645
  year: December 2010
  end-page: 654
  ident: bib0110
  article-title: Accelerating the reduction to upper Hessenberg, tridiagonal, and bidiagonal forms through hybrid GPU-based computing
  publication-title: Parallel Comput.
– year: 2014
  ident: bib0020
  publication-title: MAGMA MIC 1.3 Release: Optimizing Linear Algebra for Applications on Intel Xeon Phi Coprocessors.
– volume: 8
  start-page: 2
  year: 1987
  end-page: 13
  ident: bib0010
  article-title: The WY representation for products of Householder matrices
  publication-title: J. Sci. Stat. Comput.
– volume: 12
  start-page: 362
  year: 1991
  end-page: 373
  ident: bib0040
  article-title: Reduction of a general matrix to tridiagonal form
  publication-title: SIAM J. Mat. Anal. Appl
– reference: Rep. ORNL/TM-13344.
– volume: 28
  start-page: 196
  year: May 2014
  end-page: 209
  ident: bib0050
  article-title: A novel hybrid CPU-GPU generalized eigensolver for electronic structure calculations based on fine grained memory aware tasks
  publication-title: International Journal of High Performance Computing Applications
– reference: Geist George Howell, Gary W. and T Rowan. Error analysis of reduction to banded hessenberg form.
– reference: Intel. Math kernel library. https://software.intel.com/en-us/en-us/intel-mkl/.
– year: 2011
  ident: bib0080
  article-title: Parallel two-stage reduction to Hessenberg form using dynamic scheduling on shared-memory architectures
  publication-title: Parallel Computing
– volume: 31
  start-page: 166
  year: March 2005
  end-page: 185
  ident: bib0055
  article-title: Algorithm 841: Bhess: Gaussian reduction to a similar banded hessenberg form
  publication-title: ACM Trans. Math. Softw.
– volume: 12
  start-page: 349
  year: 1968
  end-page: 368
  ident: bib0090
  article-title: Similarity reduction of a general matrix to Hessenberg form
  publication-title: Numerische Mathematik
– year: 1995
  ident: bib0115
  publication-title: Similarity matrix reduction to banded form. manuscript
– volume: 14
  start-page: 241
  year: September 1988
  end-page: 256
  ident: bib0045
  article-title: Solution of large, dense symmetric generalized eigenvalue problems using secondary storage
  publication-title: ACM Transactions on Mathematical Software
– volume: 49
  start-page: 47
  year: 1958
  end-page: 81
  ident: bib0100
  article-title: Solution of eigenvalue problems with the LR transformation
  publication-title: Nat. Bur. Standards Appl. Math. Ser.
– volume: 26
  start-page: 602
  year: 2000
  end-page: 616
  ident: bib0015
  article-title: Algorithm 807: The SBR Toolbox—software for successive band reduction
  publication-title: ACM Transactions on Mathematical Software
– volume: 25
  start-page: 845
  year: 1999
  end-page: 860
  ident: bib0085
  article-title: Efficient eigenvalue and singular value computations on shared memory machines
  publication-title: Parallel Computing
– ident: 10.1016/j.procs.2015.05.222_bib0060
– volume: 48
  start-page: 563
  year: 2008
  ident: 10.1016/j.procs.2015.05.222_bib0075
  article-title: Blocked Algorithms for the Reduction to Hessenberg-Triangular Form Revisited
  publication-title: BIT Numerical Mathematics
  doi: 10.1007/s10543-008-0180-1
– volume: 12
  start-page: 349
  issue: 5
  year: 1968
  ident: 10.1016/j.procs.2015.05.222_bib0090
  article-title: Similarity reduction of a general matrix to Hessenberg form
  publication-title: Numerische Mathematik
  doi: 10.1007/BF02161358
– volume: 12
  start-page: 362
  year: 1991
  ident: 10.1016/j.procs.2015.05.222_bib0040
  article-title: Reduction of a general matrix to tridiagonal form
  publication-title: SIAM J. Mat. Anal. Appl
  doi: 10.1137/0612026
– year: 2014
  ident: 10.1016/j.procs.2015.05.222_bib0070
  article-title: Stanimire Tomov, and Jack Dongarra. Performance analysis and design of a Hessenberg reduction using stabilized blocked elementary transformations on new architectures
  publication-title: Computer science dept. technical report, University of Tennessee, November
– volume: 4
  start-page: 332
  year: 1961
  ident: 10.1016/j.procs.2015.05.222_bib0035
  article-title: The QR transformation, part 2
  publication-title: Computer Journal
  doi: 10.1093/comjnl/4.4.332
– volume: 31
  start-page: 166
  issue: 1
  year: 2005
  ident: 10.1016/j.procs.2015.05.222_bib0055
  article-title: Algorithm 841: Bhess: Gaussian reduction to a similar banded hessenberg form
  publication-title: ACM Trans. Math. Softw.
  doi: 10.1145/1055531.1055539
– volume: 49
  start-page: 47
  year: 1958
  ident: 10.1016/j.procs.2015.05.222_bib0100
  article-title: Solution of eigenvalue problems with the LR transformation
  publication-title: Nat. Bur. Standards Appl. Math. Ser.
– volume: 25
  start-page: 845
  issue: 7
  year: 1999
  ident: 10.1016/j.procs.2015.05.222_bib0085
  article-title: Efficient eigenvalue and singular value computations on shared memory machines
  publication-title: Parallel Computing
  doi: 10.1016/S0167-8191(99)00021-6
– volume: 26
  start-page: 602
  issue: 4
  year: 2000
  ident: 10.1016/j.procs.2015.05.222_bib0015
  article-title: Algorithm 807: The SBR Toolbox—software for successive band reduction
  publication-title: ACM Transactions on Mathematical Software
  doi: 10.1145/365723.365736
– ident: 10.1016/j.procs.2015.05.222_bib0065
– volume: 10
  start-page: 53
  year: 1991
  ident: 10.1016/j.procs.2015.05.222_bib0105
  article-title: A storage-efficient WY representation for products of Householder transformations
  publication-title: J. Sci. Stat. Comput.
  doi: 10.1137/0910005
– year: 2014
  ident: 10.1016/j.procs.2015.05.222_bib0020
  publication-title: MAGMA MIC 1.3 Release: Optimizing Linear Algebra for Applications on Intel Xeon Phi Coprocessors.
– year: 1995
  ident: 10.1016/j.procs.2015.05.222_bib0115
  publication-title: Similarity matrix reduction to banded form. manuscript
– year: 2011
  ident: 10.1016/j.procs.2015.05.222_bib0080
  article-title: Parallel two-stage reduction to Hessenberg form using dynamic scheduling on shared-memory architectures
  publication-title: Parallel Computing
– volume: 28
  start-page: 196
  issue: 2
  year: 2014
  ident: 10.1016/j.procs.2015.05.222_bib0050
  article-title: A novel hybrid CPU-GPU generalized eigensolver for electronic structure calculations based on fine grained memory aware tasks
  publication-title: International Journal of High Performance Computing Applications
  doi: 10.1177/1094342013502097
– start-page: 68
  year: 1984
  ident: 10.1016/j.procs.2015.05.222_bib0025
  publication-title: EISPACK: A package for solving matrix eigenvalue problems.
– volume: 14
  start-page: 241
  year: 1988
  ident: 10.1016/j.procs.2015.05.222_bib0045
  article-title: Solution of large, dense symmetric generalized eigenvalue problems using secondary storage
  publication-title: ACM Transactions on Mathematical Software
  doi: 10.1145/44128.44130
– volume: 8
  start-page: 2
  year: 1987
  ident: 10.1016/j.procs.2015.05.222_bib0010
  article-title: The WY representation for products of Householder matrices
  publication-title: J. Sci. Stat. Comput.
  doi: 10.1137/0908009
– year: 1999
  ident: 10.1016/j.procs.2015.05.222_bib0005
– volume: 3
  start-page: 23
  issue: 1
  year: 1960
  ident: 10.1016/j.procs.2015.05.222_bib0120
  article-title: Householder's method for the solution of the algebraic eigenproblem
  publication-title: The Computer Journal
  doi: 10.1093/comjnl/3.1.23
– volume: 36
  start-page: 645
  issue: 12
  year: 2010
  ident: 10.1016/j.procs.2015.05.222_bib0110
  article-title: Accelerating the reduction to upper Hessenberg, tridiagonal, and bidiagonal forms through hybrid GPU-based computing
  publication-title: Parallel Comput.
  doi: 10.1016/j.parco.2010.06.001
– volume: 27
  start-page: 215
  issue: 1–2
  year: 1989
  ident: 10.1016/j.procs.2015.05.222_bib0030
  article-title: Block reduction of matrices to condensed forms for eigenvalue computations
  publication-title: Journal of Computational and Applied Mathematics
– ident: 10.1016/j.procs.2015.05.222_bib0095
SSID ssj0000388917
Score 1.9758089
Snippet Many applications, ranging from big data analytics to nanostructure designs, require the solution of large dense singular value decomposition (SVD) or...
SourceID crossref
elsevier
SourceType Enrichment Source
Index Database
Publisher
StartPage 180
SubjectTerms Eigensolver
Multicore
Task-based programming
Xeon Phi
Title Performance Analysis and Optimisation of Two-sided Factorization Algorithms for Heterogeneous Platform
URI https://dx.doi.org/10.1016/j.procs.2015.05.222
Volume 51
WOSCitedRecordID wos000373939100018&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVHPJ
  databaseName: ROAD: Directory of Open Access Scholarly Resources
  customDbUrl:
  eissn: 1877-0509
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000388917
  issn: 1877-0509
  databaseCode: M~E
  dateStart: 20100101
  isFulltext: true
  titleUrlDefault: https://road.issn.org
  providerName: ISSN International Centre
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lb9NAEF6FwoEL5SkKBe2BW1jJ68eufYxoq55KDjn0Zu3LbarErlwTqh74MfxSZr3rR2gV0QOKZMWrePyYLzPj0TczCH3RGrxMFijClUwJ-FtJUl0ERIYmkExkise6HTbBz87S8_NsPpn87mphNiteluntbXb9X1UNa6BsWzr7CHX3QmEBvoPSYQtqh-0_KX4-KgXoW47Y9Ph3sA5rz96xMeLiZ0XssE49PWmH7viKzOlsdQE7zaVr1QCOCR59Baczli07X4nGih8HtW2xAeCs5afbERFT71d7ay7k0nE3LsWyHqiIp2KpHcF7dncn1kPGe11tHAVNlHDJAz_3qCovRF17fq_v5-9zFq5e0xvYlHNie844__PAmrfKvg2tM6vUTXvyHpq6AaP3jL_LQ1xZ16NsJ3aa2Kasoat73m61_ZcL7ImJHeftKm-F5FZIHiQ5CHmCnoYcXr8sPfTXkMez3XSydrBzfxtdc6uWRnjvYh4OgEZBzeIleuHfRvDMoegVmpjyNdrvJn1gb_jfoGIEKtyBCgOo8BhUuCpwDyq8BSo8gAqDILwFKtyB6i1anBwvvp0SP6GDqChOGyLCSPNYZJGGOJ6KjGnFA0FlkjIRqKBgIYdlrUQRhjqEVw2paMJkTA2jUqvoHdorq9K8RzhTKqGSFibRJmY6SSPOmSoYfKLIJPIAhd1Ty5XvXm-HqKzyHSo7QF_7g65d85bdP2edOnL_P3FxZQ4I23Xgh8ed5yN6bvdcIu8Q7TX1D_MJPVObZnlTf27h9Qf5Wa4D
linkProvider ISSN International Centre
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Performance+Analysis+and+Optimisation+of+Two-sided+Factorization+Algorithms+for+Heterogeneous+Platform&rft.jtitle=Procedia+computer+science&rft.au=Kabir%2C+Khairul&rft.au=Haidar%2C+Azzam&rft.au=Tomov%2C+Stanimire&rft.au=Dongarra%2C+Jack&rft.date=2015&rft.issn=1877-0509&rft.eissn=1877-0509&rft.volume=51&rft.spage=180&rft.epage=190&rft_id=info:doi/10.1016%2Fj.procs.2015.05.222&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_procs_2015_05_222
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1877-0509&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1877-0509&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1877-0509&client=summon