Zobrazit v EDS

PanguLU: A Scalable Regular Two-Dimensional Block-Cyclic Sparse Direct Solver on Distributed Heterogeneous Systems

Uloženo v:
Podrobná bibliografie
Název: PanguLU: A Scalable Regular Two-Dimensional Block-Cyclic Sparse Direct Solver on Distributed Heterogeneous Systems
Autoři: Xu Fu
Informace o vydavateli: Zenodo
Rok vydání: 2023
Sbírka: Zenodo
Témata: Sparse LU, Regular 2D block, Distributed platforms, Heterogeneous systems
Popis: Sparse direct solvers play a vital role in large-scale high performance scientific and engineering computations. Existing distributed sparse direct methods employ the multifrontal/supernodal patterns to aggregate columns of near-identical forms and to exploit dense basic linear algebra subprograms (BLAS) for computation. However, such data layout may bring more unevenness when the structure of the input matrix is not ideal, and using dense BLAS may waste many floating point operations on zero fill-ins, accordingly. We in this paper propose a new sparse direct solver called PanguLU. Unlike the multifrontal/supernodal layout, our work relies on simpler regular 2D blocking and stores the blocks in their sparse forms to avoid any extra fill-ins. According to the sparse patterns of the blocks, a variety of block-wise sparse BLAS methods are developed and selected for higher efficiency on local GPUs. To make PanguLU more scalable, we also adjust the mapping of blocks to processes for overall more balanced workload, and propose a synchronisation-free communication strategy considering the dependencies among different sub-tasks to reduce overall latency overhead.
Druh dokumentu: conference object
Jazyk: unknown
Relation: https://zenodo.org/records/8203219; oai:zenodo.org:8203219; https://doi.org/10.5281/zenodo.8203219
DOI: 10.5281/zenodo.8203219
Dostupnost: https://doi.org/10.5281/zenodo.8203219
https://zenodo.org/records/8203219
Rights: Creative Commons Attribution 4.0 International ; cc-by-4.0 ; https://creativecommons.org/licenses/by/4.0/legalcode
Přístupové číslo: edsbas.7F651D1C
Databáze: BASE
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://doi.org/10.5281/zenodo.8203219#
    Name: EDS - BASE (s4221598)
    Category: fullText
    Text: View record from BASE
  – Url: https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=EBSCO&SrcAuth=EBSCO&DestApp=WOS&ServiceName=TransferToWoS&DestLinkType=GeneralSearchSummary&Func=Links&author=Fu%20X
    Name: ISI
    Category: fullText
    Text: Nájsť tento článok vo Web of Science
    Icon: https://imagesrvr.epnet.com/ls/20docs.gif
    MouseOverText: Nájsť tento článok vo Web of Science
Header DbId: edsbas
DbLabel: BASE
An: edsbas.7F651D1C
RelevancyScore: 959
AccessLevel: 3
PubType: Conference
PubTypeId: conference
PreciseRelevancyScore: 958.653564453125
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: PanguLU: A Scalable Regular Two-Dimensional Block-Cyclic Sparse Direct Solver on Distributed Heterogeneous Systems
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Xu+Fu%22">Xu Fu</searchLink>
– Name: Publisher
  Label: Publisher Information
  Group: PubInfo
  Data: Zenodo
– Name: DatePubCY
  Label: Publication Year
  Group: Date
  Data: 2023
– Name: Subset
  Label: Collection
  Group: HoldingsInfo
  Data: Zenodo
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Sparse+LU%22">Sparse LU</searchLink><br /><searchLink fieldCode="DE" term="%22Regular+2D+block%22">Regular 2D block</searchLink><br /><searchLink fieldCode="DE" term="%22Distributed+platforms%22">Distributed platforms</searchLink><br /><searchLink fieldCode="DE" term="%22Heterogeneous+systems%22">Heterogeneous systems</searchLink>
– Name: Abstract
  Label: Description
  Group: Ab
  Data: Sparse direct solvers play a vital role in large-scale high performance scientific and engineering computations. Existing distributed sparse direct methods employ the multifrontal/supernodal patterns to aggregate columns of near-identical forms and to exploit dense basic linear algebra subprograms (BLAS) for computation. However, such data layout may bring more unevenness when the structure of the input matrix is not ideal, and using dense BLAS may waste many floating point operations on zero fill-ins, accordingly. We in this paper propose a new sparse direct solver called PanguLU. Unlike the multifrontal/supernodal layout, our work relies on simpler regular 2D blocking and stores the blocks in their sparse forms to avoid any extra fill-ins. According to the sparse patterns of the blocks, a variety of block-wise sparse BLAS methods are developed and selected for higher efficiency on local GPUs. To make PanguLU more scalable, we also adjust the mapping of blocks to processes for overall more balanced workload, and propose a synchronisation-free communication strategy considering the dependencies among different sub-tasks to reduce overall latency overhead.
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: conference object
– Name: Language
  Label: Language
  Group: Lang
  Data: unknown
– Name: NoteTitleSource
  Label: Relation
  Group: SrcInfo
  Data: https://zenodo.org/records/8203219; oai:zenodo.org:8203219; https://doi.org/10.5281/zenodo.8203219
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.5281/zenodo.8203219
– Name: URL
  Label: Availability
  Group: URL
  Data: https://doi.org/10.5281/zenodo.8203219<br />https://zenodo.org/records/8203219
– Name: Copyright
  Label: Rights
  Group: Cpyrght
  Data: Creative Commons Attribution 4.0 International ; cc-by-4.0 ; https://creativecommons.org/licenses/by/4.0/legalcode
– Name: AN
  Label: Accession Number
  Group: ID
  Data: edsbas.7F651D1C
PLink https://erproxy.cvtisr.sk/sfx/access?url=https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsbas&AN=edsbas.7F651D1C
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.5281/zenodo.8203219
    Languages:
      – Text: unknown
    Subjects:
      – SubjectFull: Sparse LU
        Type: general
      – SubjectFull: Regular 2D block
        Type: general
      – SubjectFull: Distributed platforms
        Type: general
      – SubjectFull: Heterogeneous systems
        Type: general
    Titles:
      – TitleFull: PanguLU: A Scalable Regular Two-Dimensional Block-Cyclic Sparse Direct Solver on Distributed Heterogeneous Systems
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Xu Fu
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2023
          Identifiers:
            – Type: issn-locals
              Value: edsbas
            – Type: issn-locals
              Value: edsbas.oa
ResultId 1