A Comparison of SuperLU Solvers on the Intel MIC Architecture

Basit Öğe Kaydı
dc.authoridÇelebi, Mustafa Serdar/0000-0003-4566-0216|Duran, Ahmet/0000-0001-9835-0006|Tuncel, Mehmet/0000-0002-6798-2876
dc.authorwosidÇelebi, Mustafa Serdar/AAJ-5557-2020
dc.authorwosidDuran, Ahmet/D-9386-2015
dc.authorwosidTuncel, Mehmet/ABD-7341-2020
dc.contributor.authorTuncel, Mehmet
dc.contributor.authorDuran, Ahmet
dc.contributor.authorCelebi, M. Serdar
dc.contributor.authorAkaydin, Bora
dc.contributor.authorTopkaya, Figen O.
dc.date.accessioned2024-07-18T20:57:00Z
dc.date.available2024-07-18T20:57:00Z
dc.date.issued2016
dc.departmentİstanbul Bilgi Üniversitesien_US
dc.description2nd International Conference on Numerical Computations - Theory and Algorithms (NUMTA) -- JUN 19-25, 2016 -- Pizzo Calabria, ITALYen_US
dc.description.abstractIn many science and engineering applications, problems may result in solving a sparse linear system AX=B. For example, SuperLU_MCDT, a linear solver, was used for the large penta-diagonal matrices for 2L) problems and hepta-diagonal matrices for 3D problems, coming from the incompressible blood flow simulation (see [1]). It is important to test the status and potential improvements of state-of-the-art solvers on new technologies. In this work, sequential, multithreaded and distributed versions of SuperLU solvers (see [2]) are examined on the Intel Xeon Phi coprocessors using offload programming model at the EURORA cluster of CINECA in Italy. We consider a portfolio of test matrices containing patterned matrices from LTEMM ([3]) and randomly located matrices. This architecture can benefit from high parallelism and large vectors. We find that the sequential Supertti benefited up to 45 % performance improvement from the offload programming depending on the sparse matrix type and the size of transferred and processed data.en_US
dc.description.sponsorshipUniv Calabria, Dept Comp Engn, Modeling, Elect & Syst Sci,Natl Inst Adv Math F Severi, Italian Natl Grp Sci Computat,Natl Res Council, Inst High Performance Comp & Networking,Univ Calabria, Int Assoc Friends,Int Assoc Math & Comp Simulat,Int Soc Global Optimizat,Soc Ind Appl Mathen_US
dc.description.sponsorshipPRACE-1IP project - EUs [RI-26I557, 2010PA1756]en_US
dc.description.sponsorshipThis research was supported by the PRACE-1IP project funded in part by the EUs 7th Framework Programme (FP7/2007-2013) under grant agreement no. RI-26I557 and the Project 2010PA1756 awarded under the 18th Call for PRACE Preparatory Access. The suggestions of the editors and two anonymous referees are also appreciated.en_US
dc.identifier.doi10.1063/1.4965394
dc.identifier.isbn978-0-7354-1438-9
dc.identifier.issn0094-243X
dc.identifier.scopus2-s2.0-84995480091en_US
dc.identifier.scopusqualityN/Aen_US
dc.identifier.urihttps://doi.org/10.1063/1.4965394
dc.identifier.urihttps://hdl.handle.net/11411/8946
dc.identifier.volume1776en_US
dc.identifier.wosWOS:000392692900083en_US
dc.identifier.wosqualityN/Aen_US
dc.indekslendigikaynakWeb of Scienceen_US
dc.indekslendigikaynakScopusen_US
dc.language.isoenen_US
dc.publisherAmer Inst Physicsen_US
dc.relation.ispartofNumerical Computations: Theory and Algorithms (Numta-2016)en_US
dc.relation.publicationcategoryKonferans Öğesi - Uluslararası - Kurum Öğretim Elemanıen_US
dc.rightsinfo:eu-repo/semantics/closedAccessen_US
dc.titleA Comparison of SuperLU Solvers on the Intel MIC Architecture
dc.typeConference Object

Dosyalar

Koleksiyon

Web of Science Indexed Publications
Scopus Indexed Publications