Architectural Implications of GNN Aggregation Programming Abstractions

Graph neural networks (GNNs) have gained significant popularity due to the powerful capability to extract useful representations from graph data. As the need for efficient GNN computation intensifies, a variety of programming abstractions designed for optimizing GNN Aggregation have emerged to facil...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE computer architecture letters Jg. 23; H. 1; S. 125 - 128
Hauptverfasser: Qi, Yingjie, Yang, Jianlei, Zhou, Ao, Qiao, Tong, Hu, Chunming
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York IEEE 01.01.2024
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
characterization
execution patterns
Graph neural networks
Graph neural networks (GNNs)
Graphical representations
Graphics processing units
Indexes
Kernel
Organizations
Programming
programming abstractions
Taxonomy
ISSN:1556-6056, 1556-6064
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Graph neural networks (GNNs) have gained significant popularity due to the powerful capability to extract useful representations from graph data. As the need for efficient GNN computation intensifies, a variety of programming abstractions designed for optimizing GNN Aggregation have emerged to facilitate acceleration. However, there is no comprehensive evaluation and analysis upon existing abstractions, thus no clear consensus on which approach is better. In this letter, we classify existing programming abstractions for GNN Aggregation by the dimension of data organization and propagation method. By constructing these abstractions on a state-of-the-art GNN library, we perform a thorough and detailed characterization study to compare their performance and efficiency, and provide several insights on future GNN acceleration based on our analysis.
AbstractList Graph neural networks (GNNs) have gained significant popularity due to the powerful capability to extract useful representations from graph data. As the need for efficient GNN computation intensifies, a variety of programming abstractions designed for optimizing GNN Aggregation have emerged to facilitate acceleration. However, there is no comprehensive evaluation and analysis upon existing abstractions, thus no clear consensus on which approach is better. In this letter, we classify existing programming abstractions for GNN Aggregation by the dimension of data organization and propagation method. By constructing these abstractions on a state-of-the-art GNN library, we perform a thorough and detailed characterization study to compare their performance and efficiency, and provide several insights on future GNN acceleration based on our analysis.
Author Qiao, Tong
Qi, Yingjie
Yang, Jianlei
Hu, Chunming
Zhou, Ao
Author_xml – sequence: 1
  givenname: Yingjie
  orcidid: 0009-0009-2785-4480
  surname: Qi
  fullname: Qi, Yingjie
  email: yingjieqi@buaa.edu.cn
  organization: School of Computer Science and Engineering, Beihang University, Beijing, China
– sequence: 2
  givenname: Jianlei
  orcidid: 0000-0001-8424-7040
  surname: Yang
  fullname: Yang, Jianlei
  email: jianlei@buaa.edu.cn
  organization: School of Computer Science and Engineering, Beihang University, Beijing, China
– sequence: 3
  givenname: Ao
  orcidid: 0009-0008-1440-1493
  surname: Zhou
  fullname: Zhou, Ao
  email: aozhou@buaa.edu.cn
  organization: School of Computer Science and Engineering, Beihang University, Beijing, China
– sequence: 4
  givenname: Tong
  orcidid: 0009-0006-6018-8414
  surname: Qiao
  fullname: Qiao, Tong
  email: by2006149@buaa.edu.cn
  organization: School of Computer Science and Engineering, Beihang University, Beijing, China
– sequence: 5
  givenname: Chunming
  orcidid: 0000-0003-3473-9703
  surname: Hu
  fullname: Hu, Chunming
  email: hucm@buaa.edu.cn
  organization: School of Computer Science and Engineering, Beihang University, Beijing, China
BookMark eNp9kL1PwzAQxS1UJNrCzsAQiTnlbCeOPUYV_ZCqwgCz5ThOcJXExU4H_nvSDyHEwHSn0_vd3XsTNOpcZxC6xzDDGMTTZp7PCBA6o5QwnMEVGuM0ZTEDlox--pTdoEkIO4CEUZ6M0SL3-sP2RvcHr5po3e4bq1VvXRciV0XL7TbK69qb-jSLXr2rvWpb29VRXoTeK33S3qLrSjXB3F3qFL0vnt_mq3jzslzP802siSB9LCqNmeClKI1KsiqtjCl0mlAFQnGTcV6UJBMcE1NSlQkoGSGgGKOFAiWKhE7R43nv3rvPgwm93LmD74aTkg5GMxggMqjYWaW9C8GbSmrbnwwMD9tGYpDHzOSQmTxmJi-ZDSD8Affetsp__Yc8nBFrjPklJ5wD5fQbRXh46A
CODEN ICALC3
CitedBy_id crossref_primary_10_1109_LCA_2025_3539371
crossref_primary_10_1109_TC_2025_3588317
Cites_doi 10.1145/3466752.3480113
10.1145/3477141
10.1145/3575693.3575723
10.1109/MICRO50266.2020.00079
10.1109/HPCA47549.2020.00012
10.1145/3447786.3456229
10.1145/3442381.3449882
ContentType Journal Article
Copyright Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2024
Copyright_xml – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2024
DBID 97E
RIA
RIE
AAYXX
CITATION
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
DOI 10.1109/LCA.2023.3326170
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005–Present
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE Electronic Library (IEL)
CrossRef
Computer and Information Systems Abstracts
Electronics & Communications Abstracts
Technology Research Database
ProQuest Computer Science Collection
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
DatabaseTitle CrossRef
Technology Research Database
Computer and Information Systems Abstracts – Academic
Electronics & Communications Abstracts
ProQuest Computer Science Collection
Computer and Information Systems Abstracts
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts Professional
DatabaseTitleList Technology Research Database
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 1556-6064
EndPage 128
ExternalDocumentID 10_1109_LCA_2023_3326170
10288038
Genre orig-research
GrantInformation_xml – fundername: National Natural Science Foundation of China
  grantid: 62072019
  funderid: 10.13039/501100001809
GroupedDBID 0R~
29I
4.4
5GY
5VS
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABQJQ
ABVLG
ACGFS
ACIWK
AENEX
AETIX
AGQYO
AGSQL
AHBIQ
AIBXA
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CS3
DU5
EBS
EJD
HZ~
IEDLZ
IFIPE
IPLJI
JAVBF
LAI
M43
O9-
OCL
PQQKQ
RIA
RIE
RNI
RNS
RZB
AAYXX
CITATION
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
ID FETCH-LOGICAL-c292t-9fc1698d9dea47f5feebc543a09a8e788bd279812ed3a790d6220a663ba0a9b43
IEDL.DBID RIE
ISICitedReferencesCount 4
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001294545700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1556-6056
IngestDate Sun Nov 30 03:53:21 EST 2025
Tue Nov 18 22:30:42 EST 2025
Sat Nov 29 01:34:54 EST 2025
Wed Aug 27 02:07:29 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 1
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c292t-9fc1698d9dea47f5feebc543a09a8e788bd279812ed3a790d6220a663ba0a9b43
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0000-0003-3473-9703
0009-0009-2785-4480
0009-0008-1440-1493
0009-0006-6018-8414
0000-0001-8424-7040
PQID 3064707902
PQPubID 75728
PageCount 4
ParticipantIDs crossref_citationtrail_10_1109_LCA_2023_3326170
crossref_primary_10_1109_LCA_2023_3326170
ieee_primary_10288038
proquest_journals_3064707902
PublicationCentury 2000
PublicationDate 2024-Jan.-June
2024-1-00
20240101
PublicationDateYYYYMMDD 2024-01-01
PublicationDate_xml – month: 01
  year: 2024
  text: 2024-Jan.-June
PublicationDecade 2020
PublicationPlace New York
PublicationPlace_xml – name: New York
PublicationTitle IEEE computer architecture letters
PublicationTitleAbbrev LCA
PublicationYear 2024
Publisher IEEE
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml – name: IEEE
– name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References ref13
Chen (ref5)
Zhang (ref7)
ref12
Wang (ref11)
ref15
Fey (ref9)
ref10
Kipf (ref1)
Ma (ref4)
Velickovic (ref14) 2017
ref8
Wang (ref2)
ref3
ref6
References_xml – ident: ref6
  doi: 10.1145/3466752.3480113
– ident: ref3
  doi: 10.1145/3477141
– ident: ref8
  doi: 10.1145/3575693.3575723
– ident: ref10
  doi: 10.1109/MICRO50266.2020.00079
– volume-title: arXiv:1609.02907
  ident: ref1
  article-title: Semi-supervised classification with graph convolutional networks
– volume-title: arXiv:1909.01315
  ident: ref11
  article-title: Deep graph library: A graph-centric, highly-performant package for graph neural networks
– ident: ref12
  doi: 10.1109/HPCA47549.2020.00012
– ident: ref13
  doi: 10.1145/3447786.3456229
– volume-title: arXiv:1903.02428
  ident: ref9
  article-title: Fast graph representation learning with PyTorch geometric
– ident: ref15
  doi: 10.1145/3442381.3449882
– start-page: 23341
  volume-title: Proc. Int. Conf. Mach. Learn.
  ident: ref2
  article-title: How powerful are spectral graph neural networks?
– start-page: 1
  volume-title: Proc. IEEE/ACM Int. Conf. Comput. Aided Des.
  ident: ref5
  article-title: fuseGNN: Accelerating graph convolutional neural network training on GPGPU
– start-page: 443
  volume-title: Proc. USENIX Conf. Usenix Annu. Tech. Conf.
  ident: ref4
  article-title: NeuGraph: Parallel deep neural network computation on large graphs
– year: 2017
  ident: ref14
  article-title: Graph attention networks
  publication-title: arXiv: 1710.10903
– start-page: 467
  volume-title: Proc. Mach. Learn. Syst.
  ident: ref7
  article-title: Understanding GNN computational graph: A coordinated computation, IO, and memory perspective
SSID ssj0046384
Score 2.3205435
Snippet Graph neural networks (GNNs) have gained significant popularity due to the powerful capability to extract useful representations from graph data. As the need...
SourceID proquest
crossref
ieee
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 125
SubjectTerms characterization
execution patterns
Graph neural networks
Graph neural networks (GNNs)
Graphical representations
Graphics processing units
Indexes
Kernel
Organizations
Programming
programming abstractions
Taxonomy
Title Architectural Implications of GNN Aggregation Programming Abstractions
URI https://ieeexplore.ieee.org/document/10288038
https://www.proquest.com/docview/3064707902
Volume 23
WOSCitedRecordID wos001294545700001&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: PRVIEE
  databaseName: IEEE Electronic Library (IEL)
  customDbUrl:
  eissn: 1556-6064
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0046384
  issn: 1556-6056
  databaseCode: RIE
  dateStart: 20020101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3NS8MwFA9uePDi_Jg4ndKDFw_ZsiRtk2MZToVRdlDZreRzCLrJPvz7TdJWJ6LgrZSXUl5e8nvfD4ArPWBcsZhD5cQFUmkxFFakUMYJdfCihQ7lY0_jNM_ZdMonVbF6qIUxxoTkM9PzjyGWrxdq411lfQ-GDBHWAI00TctirfrapU6QQgg5jhPodPTPmCTi_fEw6_kx4T1CQgPybxgUhqr8uIkDvIxa__yxA7Bf6ZFRVm78Idgx8yPQqmc0RNWRPQaj7CtS4OjvtxLIo4WNbvM8ymbO5p6Fd9GkTNd6dYAWZdK7QULdw6oNHkc3D8M7WM1OgApzvIbcqkHCmebaCJra2BojVUyJQFww4-xeqXHKHbobTUTKkU4wRsKpH1IgwSUlJ6A5X8zNKYioERgRxY2ymFotOFMqsYxZksTEJrID-jU3C1U1FvfzLV6KYGAgXjj-F57_RcX_Drj-XPFWNtX4g7bt-b1FV7K6A7r1jhXVsVsV3pzyLf8QPvtl2TnYc1-npROlC5rr5cZcgF31vn5eLS-DRH0AXcrIrw
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3NS8MwFH_oFPTitzg_e_DioVuWpG1yLOLccJYdpngraT6GoJu46d9vkrZTEQVvpbzQ8vKS3_t-AOeqw7hkEQ-lFZeQFgaHwogkLKKYWnhRQvnysftBkmXs4YEPq2J1XwujtfbJZ7rlHn0sX03lm3OVtR0YMkTYMqxElOJOWa5VX7zUipIPIkdRHFotfRGVRLw9uExbblB4ixDfgvwbCvmxKj_uYg8w3c1__toWbFSaZJCWW78NS3qyA5v1lIagOrS70E0_YwWWvv8lhTyYmuA6y4J0bK3usX8XDMuErWcLaUFaOEeIr3yY7cFd92p02Qur6QmhxBzPQ25kJ-ZMcaUFTUxktC5kRIlAXDBtLd9C4YRbfNeKiIQjFWOMhFVACoEELyjZh8ZkOtEHEFAtMCKSa2kwNUpwJmVsGDMkjoiJiya0a27msmot7iZcPOXexEA8t_zPHf_ziv9NuFiseCnbavxBu-f4_YWuZHUTjusdy6uDN8udQeWa_iF8-MuyM1jrjW4H-aCf3RzBuv0SLV0qx9CYv77pE1iV7_PH2eupl64P-TDL9g
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=Architectural+Implications+of+GNN+Aggregation+Programming+Abstractions&rft.jtitle=IEEE+computer+architecture+letters&rft.au=Qi%2C+Yingjie&rft.au=Yang%2C+Jianlei&rft.au=Zhou%2C+Ao&rft.au=Qiao%2C+Tong&rft.date=2024-01-01&rft.issn=1556-6056&rft.eissn=1556-6064&rft.volume=23&rft.issue=1&rft.spage=125&rft.epage=128&rft_id=info:doi/10.1109%2FLCA.2023.3326170&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_LCA_2023_3326170
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1556-6056&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1556-6056&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1556-6056&client=summon