Unite and conquer approach for high scale numerical computing

•We propose an approach to define numerical methods for exascale computational systems.•We propose an adapted programming model for these computing systems.•We point out the characteristics of such numerical methods.•We show that the methods defined according to the unite and conquer approach could...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of computational science Jg. 14; S. 5 - 14
Hauptverfasser: Emad, Nahid, Petiton, Serge
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.05.2016
Schlagworte:
ISSN:1877-7503, 1877-7511
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract •We propose an approach to define numerical methods for exascale computational systems.•We propose an adapted programming model for these computing systems.•We point out the characteristics of such numerical methods.•We show that the methods defined according to the unite and conquer approach could achieve our objective. The ability to exploit emerging exascale computational systems will require a careful review and redesign of core numerical algorithms and their implementations to fully exploit multiple levels of concurrency, hierarchical memory structures and heterogeneous processing units that will become available in these computational platforms. This paper presents the “unite and conquer” approach to solve linear systems of equations and eigenvalue problems for extreme scale computing. Indeed, there are two ways to optimize the execution of a restarted method on a large-scale distributed system. The first one is to optimize the number of floating point operations per restart cycle through maximizing the concurrency inside a restart cycle while minimizing latencies. The second way is to accelerate/improve the rate of convergence for a given computational scheme. The unite and conquer restarted approach focuses on decreasing the number of restart cycles by coupling either synchronously or asynchronously several restarted methods called also co-methods. In the end of a restart cycle, each co-method locally gathers available results of all collaborating co-methods and selects the best one in order to create its restarting information. Consequently this permits the global reduction of the number of cycles to convergence. The unite and conquer restarted methods are heterogeneous, fault tolerant, support asynchronous communications and present a big potential of load balancing. Due to these properties, they are well adapted to large-scale multi-level parallel architectures. We show the relevant programming paradigms that allow multi-level parallel expression of these methods and how the software engineering technology can contribute significantly in achieving high scalability. We present some experiments validating the approach for unite and conquer restarted Krylov methods on several parallel and distributed platforms.
AbstractList •We propose an approach to define numerical methods for exascale computational systems.•We propose an adapted programming model for these computing systems.•We point out the characteristics of such numerical methods.•We show that the methods defined according to the unite and conquer approach could achieve our objective. The ability to exploit emerging exascale computational systems will require a careful review and redesign of core numerical algorithms and their implementations to fully exploit multiple levels of concurrency, hierarchical memory structures and heterogeneous processing units that will become available in these computational platforms. This paper presents the “unite and conquer” approach to solve linear systems of equations and eigenvalue problems for extreme scale computing. Indeed, there are two ways to optimize the execution of a restarted method on a large-scale distributed system. The first one is to optimize the number of floating point operations per restart cycle through maximizing the concurrency inside a restart cycle while minimizing latencies. The second way is to accelerate/improve the rate of convergence for a given computational scheme. The unite and conquer restarted approach focuses on decreasing the number of restart cycles by coupling either synchronously or asynchronously several restarted methods called also co-methods. In the end of a restart cycle, each co-method locally gathers available results of all collaborating co-methods and selects the best one in order to create its restarting information. Consequently this permits the global reduction of the number of cycles to convergence. The unite and conquer restarted methods are heterogeneous, fault tolerant, support asynchronous communications and present a big potential of load balancing. Due to these properties, they are well adapted to large-scale multi-level parallel architectures. We show the relevant programming paradigms that allow multi-level parallel expression of these methods and how the software engineering technology can contribute significantly in achieving high scalability. We present some experiments validating the approach for unite and conquer restarted Krylov methods on several parallel and distributed platforms.
Author Petiton, Serge
Emad, Nahid
Author_xml – sequence: 1
  givenname: Nahid
  surname: Emad
  fullname: Emad, Nahid
  email: Nahid.Emad@uvsq.fr
  organization: Maison de la Simulation, Digitéo Labs – Bat. 565, 91191 Gif-sur-Yvette Cedex, France
– sequence: 2
  givenname: Serge
  surname: Petiton
  fullname: Petiton, Serge
  organization: Maison de la Simulation, Digitéo Labs – Bat. 565, 91191 Gif-sur-Yvette Cedex, France
BookMark eNp9kL9OwzAQhy1UJErpCzD5BRLsOIkTCQZU8U-qxEJny3HOjaPWDraLxNvjqoiBoTfc_Yb7TrrvGs2ss4DQLSU5JbS-G_PRqZAXKeeE5oTwCzSnDecZryid_WXCrtAyhJGkYk3TUjZHDxtrImBpe6yc_TyAx3KavJNqwNp5PJjtgIOSO8D2sAdvUkyb--kQjd3eoEstdwGWv3OBNs9PH6vXbP3-8rZ6XGeKsTJmNdNl2TQdyJ4WbadT7yWratnyXte6bWUHnWYMVEErqCWpuOJtAUUJpOS6ZAtUnO4q70LwoMXkzV76b0GJODoQozg6EEcHglCRHCSo-QcpE2U0zkYvze48en9CIT31ZcCLoAxYBb3xoKLonTmH_wDnZ3pa
CitedBy_id crossref_primary_10_1007_s13160_019_00359_1
crossref_primary_10_1016_j_parco_2019_102551
crossref_primary_10_1016_j_jocs_2016_04_014
Cites_doi 10.1137/0613025
10.1137/S0895479899358595
10.1137/S1064827500366082
10.1016/j.future.2005.10.003
10.1016/0168-9274(93)90103-X
10.1090/S0025-5718-1984-0736453-8
10.1007/s002110050015
10.1016/0024-3795(80)90169-X
10.1016/j.matcom.2009.07.009
10.1109/TPDS.2005.2
10.1002/1099-1506(200010/12)7:7/8<569::AID-NLA213>3.0.CO;2-8
10.1137/0907058
10.1016/j.camwa.2006.05.004
ContentType Journal Article
Copyright 2016 Elsevier B.V.
Copyright_xml – notice: 2016 Elsevier B.V.
DBID AAYXX
CITATION
DOI 10.1016/j.jocs.2016.01.007
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Sciences (General)
Business
EISSN 1877-7511
EndPage 14
ExternalDocumentID 10_1016_j_jocs_2016_01_007
S1877750316300084
GroupedDBID --K
--M
.~1
0R~
1B1
1~.
1~5
4.4
457
4G.
5VS
7-5
71M
8P~
AACTN
AAEDT
AAEDW
AAIAV
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AAXUO
AAYFN
ABBOA
ABFRF
ABJNI
ABMAC
ABXDB
ABYKQ
ACDAQ
ACGFO
ACGFS
ACNNM
ACRLP
ACZNC
ADBBV
ADEZE
ADMUD
AEBSH
AEFWE
AEKER
AENEX
AFKWA
AFTJW
AGHFR
AGUBO
AGYEJ
AHZHX
AIALX
AIEXJ
AIKHN
AITUG
AJBFU
AJOXV
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
AOUOD
AXJTR
BKOJK
BLXMC
EBS
EFJIC
EFLBG
EJD
EP3
FDB
FEDTE
FIRID
FNPLU
FYGXN
GBLVA
GBOLZ
HVGLF
HZ~
J1W
KOM
M41
MO0
N9A
O-L
O9-
OAUVE
P-8
P-9
P2P
PC.
Q38
RIG
ROL
SDF
SES
SPC
SPCBC
SSV
SSZ
T5K
UNMZH
~G-
9DU
AATTM
AAXKI
AAYWO
AAYXX
ABWVN
ACLOT
ACRPL
ACVFH
ADCNI
ADNMO
AEIPS
AEUPX
AFJKZ
AFPUW
AIGII
AIIUN
AKBMS
AKRWK
AKYEP
ANKPU
APXCP
CITATION
EFKBS
~HD
ID FETCH-LOGICAL-c334t-63f4488bead129bfd12da356a97df6f99abebf33ec215e6a057c792e24e047f43
ISICitedReferencesCount 8
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000379560000002&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1877-7503
IngestDate Tue Nov 18 21:02:45 EST 2025
Sat Nov 29 06:58:15 EST 2025
Fri Feb 23 02:31:14 EST 2024
IsPeerReviewed true
IsScholarly true
Keywords Parallel and distributed programming model
Restarted hybrid methods
Krylov methods
Extreme scale computing
Linear algebra
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c334t-63f4488bead129bfd12da356a97df6f99abebf33ec215e6a057c792e24e047f43
PageCount 10
ParticipantIDs crossref_primary_10_1016_j_jocs_2016_01_007
crossref_citationtrail_10_1016_j_jocs_2016_01_007
elsevier_sciencedirect_doi_10_1016_j_jocs_2016_01_007
PublicationCentury 2000
PublicationDate May 2016
2016-05-00
PublicationDateYYYYMMDD 2016-05-01
PublicationDate_xml – month: 05
  year: 2016
  text: May 2016
PublicationDecade 2010
PublicationTitle Journal of computational science
PublicationYear 2016
Publisher Elsevier B.V
Publisher_xml – name: Elsevier B.V
References Sorensen (bib0075) 1992; 13
Dongarra, Hittinger, Bell, Chacon, Fallout, Heroux, Hovland, Ng, Webster, Wild (bib0005) 2014
Emad, Drummond, Tsuji, Dandouna (bib0120) 2014
Delannoy, Emad, Petiton (bib0025) 2006
Thinking Machines Corporation.
Saad, Schultz (bib0065) 1986; 7
Zhang, Bergere, Petiton (bib0085) 2008
Edjlali, Emad, Petiton (bib0060) 1994
Lehoucq (bib0125) 2001; 23
Emad, Petiton, Edjlali (bib0040) 2005; 27
Delannoy (bib0020) September, 2008
He, Bergére, Petiton (bib0080) 2006; 51
Simon, Dagum (bib0105) 1993; 12
Choy, Delannoy, Emad, Petiton (bib0030) 2009
Maschhoff, Sorensen (bib0090) 1996
Arnold, Agrawal, Blackford, Dongarra, Miller, Seymour, Sagi, Shi, Vadhiyar (bib0110) June, 2002
Emad, Shahzadeh-Fazeli, Dongarra (bib0115) 2006; 22
Dookhitram, Boojhawon, Bhuruth (bib0135) 2009; 80
Shahzadeh Fazeli, Emad, Liu (bib0095) 2015
Brezinski, Redivo-Zaglia (bib0055) 1994; 67
Moriya, Nodera (bib0130) 2000; 7
Demmel (bib0045) 2013
Katagiri, Aquilanti, Petiton (bib0015) 2012
.
Saad (bib0070) 1980; 34
Saad (bib0050) 1984; 42
Emad, Delannoy, Dandouna (bib0010) 2010
Bahi, Contassot-Vivier, Couturier, Vernier (bib0035) 2005; 16
10.1016/j.jocs.2016.01.007_bib0100
Emad (10.1016/j.jocs.2016.01.007_bib0115) 2006; 22
Bahi (10.1016/j.jocs.2016.01.007_bib0035) 2005; 16
Sorensen (10.1016/j.jocs.2016.01.007_bib0075) 1992; 13
Shahzadeh Fazeli (10.1016/j.jocs.2016.01.007_bib0095) 2015
Brezinski (10.1016/j.jocs.2016.01.007_bib0055) 1994; 67
Arnold (10.1016/j.jocs.2016.01.007_bib0110) 2002
Lehoucq (10.1016/j.jocs.2016.01.007_bib0125) 2001; 23
Demmel (10.1016/j.jocs.2016.01.007_bib0045) 2013
Maschhoff (10.1016/j.jocs.2016.01.007_bib0090) 1996
Dongarra (10.1016/j.jocs.2016.01.007_bib0005) 2014
Dookhitram (10.1016/j.jocs.2016.01.007_bib0135) 2009; 80
Emad (10.1016/j.jocs.2016.01.007_bib0010) 2010
Saad (10.1016/j.jocs.2016.01.007_bib0065) 1986; 7
Simon (10.1016/j.jocs.2016.01.007_bib0105) 1993; 12
Emad (10.1016/j.jocs.2016.01.007_bib0120) 2014
Edjlali (10.1016/j.jocs.2016.01.007_bib0060) 1994
He (10.1016/j.jocs.2016.01.007_bib0080) 2006; 51
Zhang (10.1016/j.jocs.2016.01.007_bib0085) 2008
Delannoy (10.1016/j.jocs.2016.01.007_bib0020) 2008
Katagiri (10.1016/j.jocs.2016.01.007_bib0015) 2012
Delannoy (10.1016/j.jocs.2016.01.007_bib0025) 2006
Choy (10.1016/j.jocs.2016.01.007_bib0030) 2009
Emad (10.1016/j.jocs.2016.01.007_bib0040) 2005; 27
Saad (10.1016/j.jocs.2016.01.007_bib0070) 1980; 34
Moriya (10.1016/j.jocs.2016.01.007_bib0130) 2000; 7
Saad (10.1016/j.jocs.2016.01.007_bib0050) 1984; 42
References_xml – start-page: 478
  year: 1996
  end-page: 486
  ident: bib0090
  article-title: P_ARPACK: an efficient portable large scale eigenvalue package for distributed memory parallel architectures
  publication-title: Applied Parallel Computing Industrial Computation and Optimization, vol. 1184 of Lecture Notes in Computer Science
– start-page: 314
  year: 2012
  end-page: 328
  ident: bib0015
  article-title: A smart tuning strategy for restart frequency of GMRES(
  publication-title: High Performance Computing for Computational Science – VECPAR 2012, 10th International Conference, Kobe, Japan
– volume: 51
  start-page: 1647
  year: 2006
  end-page: 1662
  ident: bib0080
  article-title: A hybrid GMRES/LS-Arnoldi method to accelerate the parallel solution of linear systems
  publication-title: Comput. Math. Appl.
– volume: 12
  start-page: 431
  year: 1993
  end-page: 442
  ident: bib0105
  article-title: Experience in using SIMD and MIMD parallelism for computational fluid dynamics
  publication-title: Appl. Numer. Math.
– start-page: 286
  year: 2008
  end-page: 296
  ident: bib0085
  article-title: Grid computing: a case study in hybrid GMRES method
  publication-title: Network and Parallel Computing, vol. 5245 of Lecture Notes in Computer Science
– start-page: 451
  year: 2009
  end-page: 456
  ident: bib0030
  article-title: Federation and abstraction of heterogeneous global computing platforms with the YML framework
  publication-title: 2009 International Conference on Complex, Intelligent and Software Intensive Systems, CISIS 2009
– volume: 42
  start-page: 567
  year: 1984
  end-page: 588
  ident: bib0050
  article-title: Chebyshev acceleration techniques for solving nonsymmetric eigenvalue problems
  publication-title: Math. Comput.
– volume: 23
  start-page: 551
  year: 2001
  end-page: 562
  ident: bib0125
  article-title: Implicitly restarted Arnoldi methods and subspace iteration
  publication-title: SIAM J. Matrix Anal. Appl.
– volume: 80
  start-page: 387
  year: 2009
  end-page: 401
  ident: bib0135
  article-title: A new method for accelerating Arnoldi algorithms for large scale eigenproblems
  publication-title: Math. Comput. Simul.
– volume: 22
  start-page: 279
  year: 2006
  end-page: 290
  ident: bib0115
  article-title: An asynchronous algorithm on the NetSolve global computing system
  publication-title: Future Gener. Comput. Syst.
– volume: 7
  start-page: 569
  year: 2000
  end-page: 584
  ident: bib0130
  article-title: The DEFLATED-GMRES(
  publication-title: Numer. Linear Algebra Appl.
– start-page: 1
  year: 2015
  end-page: 20
  ident: bib0095
  article-title: A key to choose subspace size in implicitly restarted Arnoldi method
  publication-title: Numer. Algorithms
– volume: 34
  start-page: 269
  year: 1980
  end-page: 295
  ident: bib0070
  article-title: Variations on Arnoldi's method for computing eigenelements of large unsymmetric matrices
  publication-title: Linear Algebra Appl.
– volume: 7
  start-page: 856
  year: 1986
  end-page: 869
  ident: bib0065
  article-title: GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems
  publication-title: SIAM J. Sci. Stat. Comput.
– reference: Thinking Machines Corporation.
– start-page: 271
  year: 2010
  end-page: 278
  ident: bib0010
  article-title: Numerical library reuse in parallel and distributed platforms
  publication-title: High Performance Computing for Computational Science – VECPAR 2010 – 9th International Conference, Berkeley, CA, USA
– start-page: 8
  year: 2014
  ident: bib0120
  article-title: Tuning asynchronous co-methods for large-scale eigenvalue calculations
  publication-title: 16th SIAM Conference on Parallel Processing for Scientific Computing
– start-page: 585
  year: 2013
  ident: bib0045
  article-title: Communication-avoiding algorithms for linear algebra and beyond
  publication-title: 27th IEEE International Symposium on Parallel and Distributed Processing, IPDPS 2013
– year: September, 2008
  ident: bib0020
  article-title: YML: A Scientific Workflow for High Performance Computing (Ph.D. in Computer Science)
– reference: .
– year: June, 2002
  ident: bib0110
  article-title: Users’ Guide to NetSolve V1.4.1, Innovative Computing Dept., Technical Report ICL-UT-02-05
– year: 2014
  ident: bib0005
  article-title: Applied Mathematics Research for Exascale Computing, US – DOE Report
– volume: 67
  start-page: 1
  year: 1994
  end-page: 19
  ident: bib0055
  article-title: Hybrid procedures for solving linear systems
  publication-title: Numer. Math.
– volume: 27
  start-page: 253
  year: 2005
  end-page: 277
  ident: bib0040
  article-title: Multiple explicitly restarted Arnoldi method for solving large eigenproblems
  publication-title: SIAM J. Sci. Comput.
– start-page: 8
  year: 1994
  ident: bib0060
  article-title: Hybrid methods on network of heterogeneous parallel computers
  publication-title: The 14th IMACS International Symposium on Iterative Methods in Linear Algebra World Congress
– start-page: 25
  year: 2006
  end-page: 32
  ident: bib0025
  article-title: Workflow global computing with YML
  publication-title: 7th IEEE/ACM International Conference on Grid Computing
– volume: 13
  start-page: 357
  year: 1992
  end-page: 385
  ident: bib0075
  article-title: Implicit application of polynomial filters in a
  publication-title: SIAM J. Matrix Anal. Appl.
– volume: 16
  start-page: 4
  year: 2005
  end-page: 13
  ident: bib0035
  article-title: A decentralized convergence detection algorithm for asynchronous parallel iterative algorithms
  publication-title: IEEE Trans. Parallel Distrib. Syst.
– year: 2008
  ident: 10.1016/j.jocs.2016.01.007_bib0020
– year: 2014
  ident: 10.1016/j.jocs.2016.01.007_bib0005
– volume: 13
  start-page: 357
  issue: 1
  year: 1992
  ident: 10.1016/j.jocs.2016.01.007_bib0075
  article-title: Implicit application of polynomial filters in a k-step Arnoldi method
  publication-title: SIAM J. Matrix Anal. Appl.
  doi: 10.1137/0613025
– start-page: 25
  year: 2006
  ident: 10.1016/j.jocs.2016.01.007_bib0025
  article-title: Workflow global computing with YML
– volume: 23
  start-page: 551
  issue: 2
  year: 2001
  ident: 10.1016/j.jocs.2016.01.007_bib0125
  article-title: Implicitly restarted Arnoldi methods and subspace iteration
  publication-title: SIAM J. Matrix Anal. Appl.
  doi: 10.1137/S0895479899358595
– volume: 27
  start-page: 253
  issue: 1
  year: 2005
  ident: 10.1016/j.jocs.2016.01.007_bib0040
  article-title: Multiple explicitly restarted Arnoldi method for solving large eigenproblems
  publication-title: SIAM J. Sci. Comput.
  doi: 10.1137/S1064827500366082
– ident: 10.1016/j.jocs.2016.01.007_bib0100
– start-page: 585
  year: 2013
  ident: 10.1016/j.jocs.2016.01.007_bib0045
  article-title: Communication-avoiding algorithms for linear algebra and beyond
– volume: 22
  start-page: 279
  issue: 3
  year: 2006
  ident: 10.1016/j.jocs.2016.01.007_bib0115
  article-title: An asynchronous algorithm on the NetSolve global computing system
  publication-title: Future Gener. Comput. Syst.
  doi: 10.1016/j.future.2005.10.003
– start-page: 451
  year: 2009
  ident: 10.1016/j.jocs.2016.01.007_bib0030
  article-title: Federation and abstraction of heterogeneous global computing platforms with the YML framework
– volume: 12
  start-page: 431
  issue: 5
  year: 1993
  ident: 10.1016/j.jocs.2016.01.007_bib0105
  article-title: Experience in using SIMD and MIMD parallelism for computational fluid dynamics
  publication-title: Appl. Numer. Math.
  doi: 10.1016/0168-9274(93)90103-X
– volume: 42
  start-page: 567
  issue: 166
  year: 1984
  ident: 10.1016/j.jocs.2016.01.007_bib0050
  article-title: Chebyshev acceleration techniques for solving nonsymmetric eigenvalue problems
  publication-title: Math. Comput.
  doi: 10.1090/S0025-5718-1984-0736453-8
– volume: 67
  start-page: 1
  issue: 1
  year: 1994
  ident: 10.1016/j.jocs.2016.01.007_bib0055
  article-title: Hybrid procedures for solving linear systems
  publication-title: Numer. Math.
  doi: 10.1007/s002110050015
– volume: 34
  start-page: 269
  year: 1980
  ident: 10.1016/j.jocs.2016.01.007_bib0070
  article-title: Variations on Arnoldi's method for computing eigenelements of large unsymmetric matrices
  publication-title: Linear Algebra Appl.
  doi: 10.1016/0024-3795(80)90169-X
– start-page: 286
  year: 2008
  ident: 10.1016/j.jocs.2016.01.007_bib0085
  article-title: Grid computing: a case study in hybrid GMRES method
– volume: 80
  start-page: 387
  issue: 2
  year: 2009
  ident: 10.1016/j.jocs.2016.01.007_bib0135
  article-title: A new method for accelerating Arnoldi algorithms for large scale eigenproblems
  publication-title: Math. Comput. Simul.
  doi: 10.1016/j.matcom.2009.07.009
– start-page: 8
  year: 2014
  ident: 10.1016/j.jocs.2016.01.007_bib0120
  article-title: Tuning asynchronous co-methods for large-scale eigenvalue calculations
– start-page: 1
  year: 2015
  ident: 10.1016/j.jocs.2016.01.007_bib0095
  article-title: A key to choose subspace size in implicitly restarted Arnoldi method
  publication-title: Numer. Algorithms
– start-page: 478
  year: 1996
  ident: 10.1016/j.jocs.2016.01.007_bib0090
  article-title: P_ARPACK: an efficient portable large scale eigenvalue package for distributed memory parallel architectures
– volume: 16
  start-page: 4
  issue: 1
  year: 2005
  ident: 10.1016/j.jocs.2016.01.007_bib0035
  article-title: A decentralized convergence detection algorithm for asynchronous parallel iterative algorithms
  publication-title: IEEE Trans. Parallel Distrib. Syst.
  doi: 10.1109/TPDS.2005.2
– volume: 7
  start-page: 569
  issue: 7–8
  year: 2000
  ident: 10.1016/j.jocs.2016.01.007_bib0130
  article-title: The DEFLATED-GMRES(m, k) method with switching the restart frequency dynamically
  publication-title: Numer. Linear Algebra Appl.
  doi: 10.1002/1099-1506(200010/12)7:7/8<569::AID-NLA213>3.0.CO;2-8
– volume: 7
  start-page: 856
  issue: 3
  year: 1986
  ident: 10.1016/j.jocs.2016.01.007_bib0065
  article-title: GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems
  publication-title: SIAM J. Sci. Stat. Comput.
  doi: 10.1137/0907058
– start-page: 271
  year: 2010
  ident: 10.1016/j.jocs.2016.01.007_bib0010
  article-title: Numerical library reuse in parallel and distributed platforms
– volume: 51
  start-page: 1647
  issue: 11
  year: 2006
  ident: 10.1016/j.jocs.2016.01.007_bib0080
  article-title: A hybrid GMRES/LS-Arnoldi method to accelerate the parallel solution of linear systems
  publication-title: Comput. Math. Appl.
  doi: 10.1016/j.camwa.2006.05.004
– start-page: 314
  year: 2012
  ident: 10.1016/j.jocs.2016.01.007_bib0015
  article-title: A smart tuning strategy for restart frequency of GMRES(m) with hierarchical cache sizes
– start-page: 8
  year: 1994
  ident: 10.1016/j.jocs.2016.01.007_bib0060
  article-title: Hybrid methods on network of heterogeneous parallel computers
– year: 2002
  ident: 10.1016/j.jocs.2016.01.007_bib0110
SSID ssj0000388913
Score 2.0590303
Snippet •We propose an approach to define numerical methods for exascale computational systems.•We propose an adapted programming model for these computing systems.•We...
SourceID crossref
elsevier
SourceType Enrichment Source
Index Database
Publisher
StartPage 5
SubjectTerms Extreme scale computing
Krylov methods
Linear algebra
Parallel and distributed programming model
Restarted hybrid methods
Title Unite and conquer approach for high scale numerical computing
URI https://dx.doi.org/10.1016/j.jocs.2016.01.007
Volume 14
WOSCitedRecordID wos000379560000002&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: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  customDbUrl:
  eissn: 1877-7511
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000388913
  issn: 1877-7503
  databaseCode: AIEXJ
  dateStart: 20100501
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1bS8MwFA6yifgibirOG3nwQRmVdcma9nHIREWG4IS9lbRLcWN2YxfZz_fk0uymwz34Ekov2ZovPedres53ELqOI5EAD6VOFMHTRIXwHA7IO0nEpXtzqx2qEoVfWLPpt9vBqyl0OVblBFia-rNZMPxXqGEfgC1TZ7eA23YKO2AbQIcWYIf2T8ArFmnS1VKw-iOrG65CCqU-cXkMyIhyOtWfa_oqsHw6ybzYOlfVx7N1Q-M1LRn_1NOkyT-y8HhVqatrIvPfZILn4uqC681j-YxB9Blz5LfOJYtJF0xebcF36gNrVlkvEPTueoNYKqS7nlJK1dVulyWwV1yTDRjMYtF6oewjlH2EFTdUOgL5KqsFYNDy9adG-9kusEmZm0CVx7b3YJKmdHzf6p_5mZgskI3WITowI4_rGt0C2hFpEe1lSQpFVDAmeYxvjG747RHSyGNAHhvkcYY8BuSxRB4r5LFFHlvkj9H7Q6N1_-iY6hhOTAidOB5J4NXaj8AUAGeLEmg7nNQ8HrBO4iVBwCMRJYSIGFid8DgQ85gFVVGlokJZQskJyqWDVJwizEnsw2NagZdnLjOxOXHhDCYkoQb6QkvIzYYmjI10vKxg0g9_x6WEyvaaoRZO2Xh2LRvx0ExiTelCmEMbrjvb6lfO0f58jl-g3GQ0FZdoN_6adMejKzOBvgH-C3mr
linkProvider Elsevier
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=Unite+and+conquer+approach+for+high+scale+numerical+computing&rft.jtitle=Journal+of+computational+science&rft.au=Emad%2C+Nahid&rft.au=Petiton%2C+Serge&rft.date=2016-05-01&rft.issn=1877-7503&rft.volume=14&rft.spage=5&rft.epage=14&rft_id=info:doi/10.1016%2Fj.jocs.2016.01.007&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_jocs_2016_01_007
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1877-7503&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1877-7503&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1877-7503&client=summon