Domain-specific acceleration and auto-parallelization of legacy scientific code in FORTRAN 77 using source-to-source compilation

•Accelerators (GPGPUs, manycores, FPGAs) are powerful but porting code is very hard.•We automatically transform F77 code into GPU-accelerated programs using OpenCL.•Our compiler creates modern, acceleration-ready Fortran 95 from legacy FORTRAN 77.•Our compiler further creates OpenCL code with auto-p...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Computers & fluids Ročník 173; s. 1 - 5
Hlavní autoři: Vanderbauwhede, Wim, Davidson, Gavin
Médium: Journal Article
Jazyk:angličtina
Vydáno: Amsterdam Elsevier Ltd 15.09.2018
Elsevier BV
Témata:
ISSN:0045-7930, 1879-0747
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract •Accelerators (GPGPUs, manycores, FPGAs) are powerful but porting code is very hard.•We automatically transform F77 code into GPU-accelerated programs using OpenCL.•Our compiler creates modern, acceleration-ready Fortran 95 from legacy FORTRAN 77.•Our compiler further creates OpenCL code with auto-parallelized kernels.•The performance of the automatically OpenC code on GPU is as good as handported code. Massively parallel accelerators such as GPGPUs, manycores and FPGAs represent a powerful and affordable tool for scientists who look to speed up simulations of complex systems. However, porting code to such devices requires a detailed understanding of heterogeneous programming tools and effective strategies for parallelization. In this paper we present a source to source compilation approach with whole-program analysis to automatically transform single-threaded FORTRAN 77 legacy code into OpenCL-accelerated programs with parallelized kernels. The main contributions of our work are: (1) whole-source refactoring to allow any subroutine in the code to be offloaded to an accelerator. (2) Minimization of the data transfer between the host and the accelerator by eliminating redundant transfers. (3) Pragmatic auto-parallelization of the code to be offloaded to the accelerator by identification of parallelizable maps and reductions. We have validated the code transformation performance of the compiler on the NIST FORTRAN 78 test suite and several real-world codes: the Large Eddy Simulator for Urban Flows, a high-resolution turbulent flow model; the shallow water component of the ocean model Gmodel; the Linear Baroclinic Model, an atmospheric climate model and Flexpart-WRF, a particle dispersion simulator. The automatic parallelization component has been tested on as 2-D Shallow Water model (2DSW) and on the Large Eddy Simulator for Urban Flows (UFLES) and produces a complete OpenCL-enabled code base. The fully OpenCL-accelerated versions of the 2DSW and the UFLES are resp. 9x and 20x faster on GPU than the original code on CPU, in both cases this is the same performance as manually ported code.
AbstractList •Accelerators (GPGPUs, manycores, FPGAs) are powerful but porting code is very hard.•We automatically transform F77 code into GPU-accelerated programs using OpenCL.•Our compiler creates modern, acceleration-ready Fortran 95 from legacy FORTRAN 77.•Our compiler further creates OpenCL code with auto-parallelized kernels.•The performance of the automatically OpenC code on GPU is as good as handported code. Massively parallel accelerators such as GPGPUs, manycores and FPGAs represent a powerful and affordable tool for scientists who look to speed up simulations of complex systems. However, porting code to such devices requires a detailed understanding of heterogeneous programming tools and effective strategies for parallelization. In this paper we present a source to source compilation approach with whole-program analysis to automatically transform single-threaded FORTRAN 77 legacy code into OpenCL-accelerated programs with parallelized kernels. The main contributions of our work are: (1) whole-source refactoring to allow any subroutine in the code to be offloaded to an accelerator. (2) Minimization of the data transfer between the host and the accelerator by eliminating redundant transfers. (3) Pragmatic auto-parallelization of the code to be offloaded to the accelerator by identification of parallelizable maps and reductions. We have validated the code transformation performance of the compiler on the NIST FORTRAN 78 test suite and several real-world codes: the Large Eddy Simulator for Urban Flows, a high-resolution turbulent flow model; the shallow water component of the ocean model Gmodel; the Linear Baroclinic Model, an atmospheric climate model and Flexpart-WRF, a particle dispersion simulator. The automatic parallelization component has been tested on as 2-D Shallow Water model (2DSW) and on the Large Eddy Simulator for Urban Flows (UFLES) and produces a complete OpenCL-enabled code base. The fully OpenCL-accelerated versions of the 2DSW and the UFLES are resp. 9x and 20x faster on GPU than the original code on CPU, in both cases this is the same performance as manually ported code.
Massively parallel accelerators such as GPGPUs, manycores and FPGAs represent a powerful and affordable tool for scientists who look to speed up simulations of complex systems. However, porting code to such devices requires a detailed understanding of heterogeneous programming tools and effective strategies for parallelization. In this paper we present a source to source compilation approach with whole-program analysis to automatically transform single-threaded FORTRAN 77 legacy code into OpenCL-accelerated programs with parallelized kernels. The main contributions of our work are: (1) whole-source refactoring to allow any subroutine in the code to be offloaded to an accelerator. (2) Minimization of the data transfer between the host and the accelerator by eliminating redundant transfers. (3) Pragmatic auto-parallelization of the code to be offloaded to the accelerator by identification of parallelizable maps and reductions. We have validated the code transformation performance of the compiler on the NIST FORTRAN 78 test suite and several real-world codes: the Large Eddy Simulator for Urban Flows, a high-resolution turbulent flow model; the shallow water component of the ocean model Gmodel; the Linear Baroclinic Model, an atmospheric climate model and Flexpart-WRF, a particle dispersion simulator. The automatic parallelization component has been tested on as 2-D Shallow Water model (2DSW) and on the Large Eddy Simulator for Urban Flows (UFLES) and produces a complete OpenCL-enabled code base. The fully OpenCL-accelerated versions of the 2DSW and the UFLES are resp. 9x and 20x faster on GPU than the original code on CPU, in both cases this is the same performance as manually ported code.
Author Vanderbauwhede, Wim
Davidson, Gavin
Author_xml – sequence: 1
  givenname: Wim
  orcidid: 0000-0001-6768-0037
  surname: Vanderbauwhede
  fullname: Vanderbauwhede, Wim
  email: wim.vanderbauwhede@glasgow.ac.uk, wim@dcs.gla.ac.uk
– sequence: 2
  givenname: Gavin
  surname: Davidson
  fullname: Davidson, Gavin
BookMark eNqNkE9P3DAQxa0KpC5_PgOWek4YJ9nYOfSwoqUgIZAQnC3HGSOvvHZqJ5XoqR-93g3qgUs5jUee9-bN74Qc-eCRkAsGJQPWXm5LHXajcbMdygqYKKEtAdafyIoJ3hXAG35EVgDNuuBdDZ_JSUpbyH1dNSvy51vYKeuLNKK2xmqqtEaHUU02eKr8QNU8hWJUUTmHzv5ePoKhDl-UfqVJW_TTQarDgNR6ev3w-PS4uaec0zlZ_0JTmKPGIvssL7pPbN3B6owcG-USnr_VU_J8_f3p6qa4e_hxe7W5K3TdVVPRgWi5aMGITvSmNpXhjR4aVJ3Kx_atYahqrvu-QWFA9ZwhVGrIN_Zr061FfUq-LL5jDD9nTJPc5iw-r5QVY01VgeD7qa_LlI4hpYhGajsdck5RWScZyD10uZX_oMs9dAmtzDmynr_Tj9HuVHz9gHKzKDFD-GUxygNZjYONqCc5BPtfj79FfaYi
CitedBy_id crossref_primary_10_1007_s10766_018_0572_z
crossref_primary_10_1007_s11227_021_03839_9
crossref_primary_10_1109_ACCESS_2020_3024098
crossref_primary_10_1007_s11227_022_04925_2
crossref_primary_10_1155_2019_7348013
Cites_doi 10.5194/gmd-6-1889-2013
10.1175/1520-0442(2003)16<1121:AMLBMC>2.0.CO;2
10.1109/MCSE.2010.69
10.1175/1520-0485-32.9.2509
10.1145/2541348.2541356
10.1175/2011JAMC2567.1
ContentType Journal Article
Copyright 2018 The Authors
Copyright Elsevier BV Sep 15, 2018
Copyright_xml – notice: 2018 The Authors
– notice: Copyright Elsevier BV Sep 15, 2018
DBID 6I.
AAFTH
AAYXX
CITATION
7SC
7TB
7U5
8FD
FR3
H8D
JQ2
KR7
L7M
L~C
L~D
DOI 10.1016/j.compfluid.2018.06.005
DatabaseName ScienceDirect Open Access Titles
Elsevier:ScienceDirect:Open Access
CrossRef
Computer and Information Systems Abstracts
Mechanical & Transportation Engineering Abstracts
Solid State and Superconductivity Abstracts
Technology Research Database
Engineering Research Database
Aerospace Database
ProQuest Computer Science Collection
Civil Engineering Abstracts
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
DatabaseTitle CrossRef
Aerospace Database
Civil Engineering Abstracts
Technology Research Database
Computer and Information Systems Abstracts – Academic
Mechanical & Transportation Engineering Abstracts
ProQuest Computer Science Collection
Computer and Information Systems Abstracts
Solid State and Superconductivity Abstracts
Engineering Research Database
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts Professional
DatabaseTitleList
Aerospace Database
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISSN 1879-0747
EndPage 5
ExternalDocumentID 10_1016_j_compfluid_2018_06_005
S0045793018302950
GroupedDBID --K
--M
-~X
.DC
.~1
0R~
1B1
1~.
1~5
4.4
457
4G.
5GY
5VS
6I.
7-5
71M
8P~
9JN
AACTN
AAEDT
AAEDW
AAFTH
AAIAV
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AAXUO
ABAOU
ABJNI
ABMAC
ABYKQ
ACAZW
ACDAQ
ACGFS
ACIWK
ACRLP
ADBBV
ADEZE
ADGUI
ADTZH
AEBSH
AECPX
AEKER
AENEX
AFKWA
AFTJW
AGHFR
AGUBO
AGYEJ
AHHHB
AHJVU
AIEXJ
AIGVJ
AIKHN
AITUG
AJOXV
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
ARUGR
AXJTR
BJAXD
BKOJK
BLXMC
CS3
DU5
EBS
EFJIC
EFLBG
EJD
EO8
EO9
EP2
EP3
FDB
FIRID
FNPLU
FYGXN
G-Q
GBLVA
IHE
J1W
JJJVA
KOM
LG9
LY7
M41
MHUIS
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
PQQKQ
Q38
ROL
RPZ
SDF
SDG
SDP
SES
SPC
SPCBC
SPD
SST
SSW
SSZ
T5K
TN5
XPP
ZMT
~G-
29F
6TJ
9DU
AAQXK
AATTM
AAXKI
AAYWO
AAYXX
ABDPE
ABEFU
ABFNM
ABWVN
ABXDB
ACKIV
ACLOT
ACNNM
ACRPL
ACVFH
ADCNI
ADIYS
ADMUD
ADNMO
AEIPS
AEUPX
AFFNX
AFJKZ
AFPUW
AGQPQ
AI.
AIGII
AIIUN
AKBMS
AKRWK
AKYEP
ANKPU
APXCP
ASPBG
AVWKF
AZFZN
CITATION
EFKBS
FEDTE
FGOYB
G-2
HLZ
HVGLF
HZ~
R2-
SBC
SET
SEW
T9H
VH1
WUQ
~HD
7SC
7TB
7U5
8FD
FR3
H8D
JQ2
KR7
L7M
L~C
L~D
ID FETCH-LOGICAL-c392t-90867860f898bf3f2f74cd4ea9a005b6f1ea37cbb4e8f0ab71e02ad324b5f9583
ISICitedReferencesCount 8
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000443788400001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0045-7930
IngestDate Sun Nov 09 06:53:22 EST 2025
Sat Nov 29 07:27:11 EST 2025
Tue Nov 18 22:23:51 EST 2025
Fri Feb 23 02:46:21 EST 2024
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Keywords Source-to-source compilation
Auto-parallelization
OpenCL
Acceleration
GPGPU
Fortran
Language English
License This is an open access article under the CC BY license.
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c392t-90867860f898bf3f2f74cd4ea9a005b6f1ea37cbb4e8f0ab71e02ad324b5f9583
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0000-0001-6768-0037
OpenAccessLink https://dx.doi.org/10.1016/j.compfluid.2018.06.005
PQID 2114220878
PQPubID 2045265
PageCount 5
ParticipantIDs proquest_journals_2114220878
crossref_citationtrail_10_1016_j_compfluid_2018_06_005
crossref_primary_10_1016_j_compfluid_2018_06_005
elsevier_sciencedirect_doi_10_1016_j_compfluid_2018_06_005
PublicationCentury 2000
PublicationDate 2018-09-15
PublicationDateYYYYMMDD 2018-09-15
PublicationDate_xml – month: 09
  year: 2018
  text: 2018-09-15
  day: 15
PublicationDecade 2010
PublicationPlace Amsterdam
PublicationPlace_xml – name: Amsterdam
PublicationTitle Computers & fluids
PublicationYear 2018
Publisher Elsevier Ltd
Elsevier BV
Publisher_xml – name: Elsevier Ltd
– name: Elsevier BV
References Tinetti, Méndez (bib0001) 2012; 31
Overbey, Xanthos, Johnson, Foote (bib0006) 2005
Kämpf (bib0008) 2009
Watanabe, Jin (bib0011) 2003; 16
Vanderbauwhede, Takemi (bib0003) 2015
Stone, Gohara, Shi (bib0002) 2010; 12
Orchard, Rice (bib0005) 2013
Amini, Ancourt, Coelho, Creusillet, Guelton, Irigoin (bib0007) 2011
Burgers, Balmaseda, Vossepoel, van Oldenborgh, Van Leeuwen (bib0009) 2002; 32
Liao, Quinlan, Panas, De Supinski (bib0004) 2010
Brioude, Arnold, Stohl, Cassiani, Morton, Seibert (bib0010) 2013; 6
Nakayama, Takemi, Nagai (bib0012) 2011; 50
Liao (10.1016/j.compfluid.2018.06.005_bib0004) 2010
Kämpf (10.1016/j.compfluid.2018.06.005_bib0008) 2009
Stone (10.1016/j.compfluid.2018.06.005_bib0002) 2010; 12
Nakayama (10.1016/j.compfluid.2018.06.005_bib0012) 2011; 50
Brioude (10.1016/j.compfluid.2018.06.005_bib0010) 2013; 6
Overbey (10.1016/j.compfluid.2018.06.005_bib0006) 2005
Burgers (10.1016/j.compfluid.2018.06.005_bib0009) 2002; 32
Vanderbauwhede (10.1016/j.compfluid.2018.06.005_bib0003) 2015
Amini (10.1016/j.compfluid.2018.06.005_bib0007) 2011
Tinetti (10.1016/j.compfluid.2018.06.005_bib0001) 2012; 31
Orchard (10.1016/j.compfluid.2018.06.005_bib0005) 2013
Watanabe (10.1016/j.compfluid.2018.06.005_bib0011) 2003; 16
References_xml – volume: 6
  start-page: 1889
  year: 2013
  end-page: 1904
  ident: bib0010
  article-title: The lagrangian particle dispersion model flexpart-wrf version 3.1
  publication-title: Geosci Model Dev
– volume: 12
  start-page: 66
  year: 2010
  end-page: 73
  ident: bib0002
  article-title: OpenCL: a parallel programming standard for heterogeneous computing systems
  publication-title: Comput Sci Eng
– start-page: 29
  year: 2013
  end-page: 32
  ident: bib0005
  article-title: Upgrading fortran source code using automatic refactoring
  publication-title: Proceedings of the 2013 ACM workshop on workshop on refactoring tools, WRT’13
– volume: 16
  start-page: 1121
  year: 2003
  end-page: 1139
  ident: bib0011
  article-title: A moist linear baroclinic model: coupled dynamical–convective response to el niño
  publication-title: J Clim
– start-page: 436
  year: 2015
  end-page: 443
  ident: bib0003
  article-title: Twinned buffering: a simple and highly effective scheme for parallelization of successive over-relaxation on gpus and other accelerators
  publication-title: High performance computing & simulation (HPCS), 2015 int’l conf on
– start-page: 15
  year: 2010
  end-page: 28
  ident: bib0004
  article-title: A rose-based openmp 3.0 research compiler supporting multiple runtime libraries
  publication-title: International workshop on OpenMP
– year: 2009
  ident: bib0008
  article-title: Ocean modelling for beginners: using open-source software
– start-page: 1
  year: 2011
  end-page: 6
  ident: bib0007
  article-title: Pips is not (just) polyhedral software adding gpu code generation in pips
  publication-title: First international workshop on polyhedral compilation techniques (IMPACT 2011) in conjunction with CGO 2011
– volume: 32
  start-page: 2509
  year: 2002
  end-page: 2519
  ident: bib0009
  article-title: Balanced ocean-data assimilation near the equator
  publication-title: J Phys Oceanogr
– volume: 31
  start-page: 5
  year: 2012
  end-page: 22
  ident: bib0001
  article-title: Fortran legacy software: source code update and possible parallelisation issues
  publication-title: ACM SIGPLAN fortran forum
– start-page: 37
  year: 2005
  end-page: 39
  ident: bib0006
  article-title: Refactorings for fortran and high-performance computing
  publication-title: Proceedings of the second international workshop on Software engineering for high performance computing system applications
– volume: 50
  start-page: 1692
  year: 2011
  end-page: 1712
  ident: bib0012
  article-title: Les analysis of the aerodynamic surface properties for turbulent flows over building arrays with various geometries
  publication-title: J Appl Meteorol Climatol
– year: 2009
  ident: 10.1016/j.compfluid.2018.06.005_bib0008
– volume: 6
  start-page: 1889
  issue: 6
  year: 2013
  ident: 10.1016/j.compfluid.2018.06.005_bib0010
  article-title: The lagrangian particle dispersion model flexpart-wrf version 3.1
  publication-title: Geosci Model Dev
  doi: 10.5194/gmd-6-1889-2013
– volume: 16
  start-page: 1121
  issue: 8
  year: 2003
  ident: 10.1016/j.compfluid.2018.06.005_bib0011
  article-title: A moist linear baroclinic model: coupled dynamical–convective response to el niño
  publication-title: J Clim
  doi: 10.1175/1520-0442(2003)16<1121:AMLBMC>2.0.CO;2
– volume: 12
  start-page: 66
  issue: 3
  year: 2010
  ident: 10.1016/j.compfluid.2018.06.005_bib0002
  article-title: OpenCL: a parallel programming standard for heterogeneous computing systems
  publication-title: Comput Sci Eng
  doi: 10.1109/MCSE.2010.69
– start-page: 37
  year: 2005
  ident: 10.1016/j.compfluid.2018.06.005_bib0006
  article-title: Refactorings for fortran and high-performance computing
– volume: 32
  start-page: 2509
  issue: 9
  year: 2002
  ident: 10.1016/j.compfluid.2018.06.005_bib0009
  article-title: Balanced ocean-data assimilation near the equator
  publication-title: J Phys Oceanogr
  doi: 10.1175/1520-0485-32.9.2509
– start-page: 29
  year: 2013
  ident: 10.1016/j.compfluid.2018.06.005_bib0005
  article-title: Upgrading fortran source code using automatic refactoring
  doi: 10.1145/2541348.2541356
– start-page: 1
  year: 2011
  ident: 10.1016/j.compfluid.2018.06.005_bib0007
  article-title: Pips is not (just) polyhedral software adding gpu code generation in pips
– start-page: 15
  year: 2010
  ident: 10.1016/j.compfluid.2018.06.005_bib0004
  article-title: A rose-based openmp 3.0 research compiler supporting multiple runtime libraries
– start-page: 436
  year: 2015
  ident: 10.1016/j.compfluid.2018.06.005_bib0003
  article-title: Twinned buffering: a simple and highly effective scheme for parallelization of successive over-relaxation on gpus and other accelerators
– volume: 50
  start-page: 1692
  issue: 8
  year: 2011
  ident: 10.1016/j.compfluid.2018.06.005_bib0012
  article-title: Les analysis of the aerodynamic surface properties for turbulent flows over building arrays with various geometries
  publication-title: J Appl Meteorol Climatol
  doi: 10.1175/2011JAMC2567.1
– volume: 31
  start-page: 5
  year: 2012
  ident: 10.1016/j.compfluid.2018.06.005_bib0001
  article-title: Fortran legacy software: source code update and possible parallelisation issues
SSID ssj0004324
Score 2.2818644
Snippet •Accelerators (GPGPUs, manycores, FPGAs) are powerful but porting code is very hard.•We automatically transform F77 code into GPU-accelerated programs using...
Massively parallel accelerators such as GPGPUs, manycores and FPGAs represent a powerful and affordable tool for scientists who look to speed up simulations of...
SourceID proquest
crossref
elsevier
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 1
SubjectTerms Acceleration
Auto-parallelization
Climate models
Complex systems
Computational fluid dynamics
Computer peripherals
Computer simulation
Data transfer (computers)
FORTRAN
GPGPU
Ocean models
OpenCL
Particle accelerators
Physical simulation
Program verification (computers)
Shallow water
Software
Software testing
Source-to-source compilation
Turbulent flow
Two dimensional models
Vortices
Title Domain-specific acceleration and auto-parallelization of legacy scientific code in FORTRAN 77 using source-to-source compilation
URI https://dx.doi.org/10.1016/j.compfluid.2018.06.005
https://www.proquest.com/docview/2114220878
Volume 173
WOSCitedRecordID wos000443788400001&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: 1879-0747
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0004324
  issn: 0045-7930
  databaseCode: AIEXJ
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lj9MwELZKlwMcEE-xsCAfuFWR8nLscKtgeR0qhAr0FtmuDVll02rbdPe4En-c8SNNqIrKHrhEViRbib8v4_FkvjFCr6KMSMLTFKwfT4OUZ_DN5VIH4IurXBvlgHaHTdDJhM1m-efB4FerhdlUtK7Z1VW-_K9Qwz0A20hnbwD3dlC4AW0AHa4AO1z_Cfi3i3PY7QdGQmnSgEZcSlhaPNC2NGuzXgSm5HdVqcrLMI3PWKkf5ux3J5G0XY3e3QREYKc4_TKejCgdNTa24EL-AYzjWjYzvaw6lNvaB_7MiJVlmK6acr714b9ZXY3gzeVP5Wr_fi_Pu8D5pmzFYO-hXfejE5FNpXD6zNbipqYkpv_30lpcmvRsZtRbfMleq-4CDGcGlKV9VJOS5-quhqRbyNqf9zvr2zbrsE1oOyu2AxVmoMKm9pFb6CimJGdDdDT-eDr71Mlrk9gV8_av8kea4N5n-puTs7PcWx9meh_d85sPPHakeYAGqn6I7vZKUj5C1zv0wX36YIAM76MPXmjs6IM7-mBDH1zW2NMHU4otffAufXCPPo_R13en0zcfAn9KRyDBt14HOWyKKctCzXImdKJjTVM5TxXPOcyEyHSkeEKlEKliOuSCRiqM-RzmVBCdE5Y8QcN6UaunCAsho1iFWZgwkcZcC6q0ZiqSEjbOGSHHKGuntZC-hL05SaUqDkB7jMJtx6Wr4nK4y-sWt8I7o87JLICVhzuftEgX3jSsitjK1kNG2bObP85zdKf7vE7QcH3RqBfottysy9XFS8_Y30Kgv6s
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=Domain-specific+acceleration+and+auto-parallelization+of+legacy+scientific+code+in+FORTRAN+77+using+source-to-source+compilation&rft.jtitle=Computers+%26+fluids&rft.au=Vanderbauwhede%2C+Wim&rft.au=Davidson%2C+Gavin&rft.date=2018-09-15&rft.issn=0045-7930&rft.volume=173&rft.spage=1&rft.epage=5&rft_id=info:doi/10.1016%2Fj.compfluid.2018.06.005&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_compfluid_2018_06_005
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0045-7930&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0045-7930&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0045-7930&client=summon