Mixed- Precision Algorithm in Electromagnetic Finite Element Method

Single-precision algorithms are known to offer superior computational efficiency compared to double-precision algorithms. However, achieving the same level of computational accuracy as double-precision algorithms in single-precision arithmetic can be challenging. Mixed-precision algorithms have emer...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2023 International Applied Computational Electromagnetics Society Symposium (ACES-China) s. 1 - 3
Hlavní autoři: Zhi-ming, Li, Hao-xiang, Wu, Sheng, Zuo, Xi, Chen
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: Applied Computational Electromagnetics Society (ACES) 15.08.2023
Témata:
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Single-precision algorithms are known to offer superior computational efficiency compared to double-precision algorithms. However, achieving the same level of computational accuracy as double-precision algorithms in single-precision arithmetic can be challenging. Mixed-precision algorithms have emerged as a promising solution to this problem. In this study, we employ a mixed-precision algorithm to solve the matrix equation Ax=b generated by the electromagnetic finite element method. Specifically, single-precision arithmetic is utilized in the matrix factorization phase, whereas double-precision arithmetic is applied in the residual calculation and iterative refinement phase. Our experimental results demonstrate that the mixed-precision algorithm achieves computational efficiency that is approximately 1.5 times higher than that of the double-precision algorithm, and occupies 40% less memory. Additionally, the computational accuracy of the mixed-precision algorithm is found to meet the requirements of double-precision arithmetic.
AbstractList Single-precision algorithms are known to offer superior computational efficiency compared to double-precision algorithms. However, achieving the same level of computational accuracy as double-precision algorithms in single-precision arithmetic can be challenging. Mixed-precision algorithms have emerged as a promising solution to this problem. In this study, we employ a mixed-precision algorithm to solve the matrix equation Ax=b generated by the electromagnetic finite element method. Specifically, single-precision arithmetic is utilized in the matrix factorization phase, whereas double-precision arithmetic is applied in the residual calculation and iterative refinement phase. Our experimental results demonstrate that the mixed-precision algorithm achieves computational efficiency that is approximately 1.5 times higher than that of the double-precision algorithm, and occupies 40% less memory. Additionally, the computational accuracy of the mixed-precision algorithm is found to meet the requirements of double-precision arithmetic.
Author Hao-xiang, Wu
Xi, Chen
Sheng, Zuo
Zhi-ming, Li
Author_xml – sequence: 1
  givenname: Li
  surname: Zhi-ming
  fullname: Zhi-ming, Li
  email: 18819268955@163.com
  organization: School of Electronic Engineering Xidian University,Xi'an,China
– sequence: 2
  givenname: Wu
  surname: Hao-xiang
  fullname: Hao-xiang, Wu
  email: xduwhx@outlook.com
  organization: School of Electronic Engineering Xidian University,Xi'an,China
– sequence: 3
  givenname: Zuo
  surname: Sheng
  fullname: Sheng, Zuo
  email: zuosheng0503@163.com
  organization: School of Electronic Engineering Xidian University,Xi'an,China
– sequence: 4
  givenname: Chen
  surname: Xi
  fullname: Xi, Chen
  email: chenxi@mail.xidian.edu.cn
  organization: School of Electronic Engineering Xidian University,Xi'an,China
BookMark eNo1j01Lw0AUAFfQg9b-Aw978pa4X9nNO4aQqtCi0Houm81L8yDZSLoH_fci2tPAHAbmjl3HOSJjj1LkSoOEp6pu9lk9UPRWqBJyJZTOpVAGSuOu2BpcKZ3WhQBbuFtW7-gLu4y_LxjoTHPk1XiaF0rDxCnyZsSQlnnyp4iJAt9QpIS_esKY-A7TMHf37Kb34xnX_1yxj01zqF-y7dvza11tM5ISUhaUdU54dF0rAbzXxpfBFtYH17bQOwDTmU6CUr3EUOoAoUAtdNtaj6iNXrGHvy4h4vFzockv38fLm_4BsVRLwQ
ContentType Conference Proceeding
DBID 6IE
6IL
CBEJK
RIE
RIL
DOI 10.23919/ACES-China60289.2023.10249847
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Xplore POP ALL
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP All) 1998-Present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
EISBN 9781733509657
1733509658
EndPage 3
ExternalDocumentID 10249847
Genre orig-research
GrantInformation_xml – fundername: Fundamental Research Funds for the Central Universities
  grantid: QTZX22160
  funderid: 10.13039/501100012226
– fundername: Key Research and Development Program of Shaanxi
  grantid: 2023-ZDLGY-42,2022ZDLGY02-05,2021GXLH-02,2023-ZDLGY-09
  funderid: 10.13039/501100013317
GroupedDBID 6IE
6IL
CBEJK
RIE
RIL
ID FETCH-LOGICAL-i119t-c26770ae7db199aa34a8c656ac7bb9f7994d4d1922f1ec83c9c5e303bb6aee343
IEDL.DBID RIE
IngestDate Wed Sep 27 05:40:30 EDT 2023
IsPeerReviewed false
IsScholarly false
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i119t-c26770ae7db199aa34a8c656ac7bb9f7994d4d1922f1ec83c9c5e303bb6aee343
PageCount 3
ParticipantIDs ieee_primary_10249847
PublicationCentury 2000
PublicationDate 2023-Aug.-15
PublicationDateYYYYMMDD 2023-08-15
PublicationDate_xml – month: 08
  year: 2023
  text: 2023-Aug.-15
  day: 15
PublicationDecade 2020
PublicationTitle 2023 International Applied Computational Electromagnetics Society Symposium (ACES-China)
PublicationTitleAbbrev ACES-CHINA
PublicationYear 2023
Publisher Applied Computational Electromagnetics Society (ACES)
Publisher_xml – name: Applied Computational Electromagnetics Society (ACES)
Score 1.8412558
Snippet Single-precision algorithms are known to offer superior computational efficiency compared to double-precision algorithms. However, achieving the same level of...
SourceID ieee
SourceType Publisher
StartPage 1
SubjectTerms Approximation algorithms
Computational efficiency
Computational modeling
Electromagnetic Finite Element method
Finite element analysis
Iterative methods
Mathematical models
Memory management
mixed-precision algorithm
Sparse matrix direct solver
Title Mixed- Precision Algorithm in Electromagnetic Finite Element Method
URI https://ieeexplore.ieee.org/document/10249847
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEB5sEfGkYsU3OYi31KbJbpJjKS0ebCn4oLeSx2xd0K3UVvz5JmmrePDgLUxIQhImM5PJlw_gytnCGykyajgzVBTOU6MyQzNWKKdt4bL0genTnRwO1XisR2uwesLCIGJ6fIbNWEy5fD9zy3hVFjQ8BAvhOK1BTcp8Bdbagev0nFkzfdPp9u5p4p3OYwKtGanBm5tGv-hTkvXo7_1z3H1o_ODwyOjbwhzAFlaH0B2Un-hpqFjT45DOy3QWYvznV1JWpLfitXk10yriE0m_jF5lFMdByCAxRjfgsd976N7SNRUCLRnTC-rauZQtg9JbprUxXBjlgitmnLRWF1Jr4YUP3lq7YOgUd9plGKyTtblB5IIfQb2aVXgMhOXShV5Q2RYXyhvFc6asDHppJWJLnEAjrsDkbfXbxWQz-dM_5GewG9c53rOy7Bzqi_kSL2DbfSzK9_ll2qMvtIOVHw
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LTwIxEJ4oGvWkRoxv92C8FSnb3bZHQiAYgZCIhhvpYxY3kcUgGH--bQGNBw_emmnapm2mM9Pp1w_gxujMKs4SomKqCMuMJUokiiQ0E0bqzCThA9PnDu_1xHAo-yuwesDCIGJ4fIYVXwy5fDs1C39V5jTcBQvuON2ELU-dtYJr7cBteNAsqbyrN5qPJDBPpz6FVvHk4JV1s18EKsF-tPb_OfIBlH-QeFH_28YcwgYWR9Do5p9oiatYEeRE9dfx1EX5L5MoL6LmktlmosaFRyhGrdz7lV7sB4m6gTO6DE-t5qDRJisyBJJTKufE1FLOqwq51VRKpWKmhHHOmDJca5lxKZll1vlrtYyiEbGRJkFnn7ROFWLM4mMoFdMCTyCiKTeuFxS6GjNhlYhTKjR3mqk5YpWdQtmvwOht-d_FaD35sz_k17DbHnQ7o8597-Ec9vya-1tXmlxAaT5b4CVsm495_j67Cvv1BaqSmGg
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%3Abook&rft.genre=proceeding&rft.title=2023+International+Applied+Computational+Electromagnetics+Society+Symposium+%28ACES-China%29&rft.atitle=Mixed-+Precision+Algorithm+in+Electromagnetic+Finite+Element+Method&rft.au=Zhi-ming%2C+Li&rft.au=Hao-xiang%2C+Wu&rft.au=Sheng%2C+Zuo&rft.au=Xi%2C+Chen&rft.date=2023-08-15&rft.pub=Applied+Computational+Electromagnetics+Society+%28ACES%29&rft.spage=1&rft.epage=3&rft_id=info:doi/10.23919%2FACES-China60289.2023.10249847&rft.externalDocID=10249847