Matrices over runtime systems at exascale

Emmanuel Agullo; George Bosilca; Bérenger Bramas; Cedric Castagnede; Olivier Coulaud; Eric F. Darve; Jack Dongarra; Mathieu Faverge; Nathalie Furmento; Luc Giraud; Xavier Lacoste; Julien Langou; Hatem Ltaief; Matthias Messner; Raymond Namyst; Pierre Ramet; Toru Takahashi; Samuel Thibault; Stanimire Z. Tomov; Ichitaro Yamazaki

doi:10.1109/SC.Companion.2012.167

Matrices over runtime systems at exascale

Emmanuel Agullo, George Bosilca, Bérenger Bramas, Cedric Castagnede, Olivier Coulaud, Eric F. Darve, Jack Dongarra, Mathieu Faverge, Nathalie Furmento, Luc Giraud, Xavier Lacoste, Julien Langou, Hatem Ltaief, Matthias Messner, Raymond Namyst, Pierre Ramet, Toru Takahashi, Samuel Thibault, Stanimire Z. Tomov, Ichitaro Yamazaki

Supercomputing Laboratory

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Scopus citations

Abstract

The goal of Matrices Over Runtime Systems at Exascale (MORSE) project is to design dense and sparse linear algebra methods that achieve the fastest possible time to an accurate solution on large-scale multicore systems with GPU accelerators, using all the processing power that future high end systems can make available. In this poster, we propose a framework for describing linear algebra algorithms at a high level of abstraction and delegating the actual execution to a runtime system in order to design software whose performance is portable accross architectures. We illustrate our methodology on three classes of problems: dense linear algebra, sparse direct methods and fast multipole methods. The resulting codes have been incorporated into Magma, Pastix and ScalFMM solvers, respectively. © 2012 IEEE.

Original language	English (US)
Title of host publication	2012 SC Companion: High Performance Computing, Networking Storage and Analysis
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	1330-1332
Number of pages	3
ISBN (Print)	9780769549569
DOIs	https://doi.org/10.1109/SC.Companion.2012.167
State	Published - Nov 2012

Access to Document

10.1109/SC.Companion.2012.167

Cite this

Agullo, E., Bosilca, G., Bramas, B., Castagnede, C., Coulaud, O., Darve, E. F., Dongarra, J., Faverge, M., Furmento, N., Giraud, L., Lacoste, X., Langou, J., Ltaief, H., Messner, M., Namyst, R., Ramet, P., Takahashi, T., Thibault, S., Tomov, S. Z., & Yamazaki, I. (2012). Matrices over runtime systems at exascale. In 2012 SC Companion: High Performance Computing, Networking Storage and Analysis (pp. 1330-1332). Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/SC.Companion.2012.167

@inproceedings{20318c820a354070944c694101b19f01,

title = "Matrices over runtime systems at exascale",

abstract = "The goal of Matrices Over Runtime Systems at Exascale (MORSE) project is to design dense and sparse linear algebra methods that achieve the fastest possible time to an accurate solution on large-scale multicore systems with GPU accelerators, using all the processing power that future high end systems can make available. In this poster, we propose a framework for describing linear algebra algorithms at a high level of abstraction and delegating the actual execution to a runtime system in order to design software whose performance is portable accross architectures. We illustrate our methodology on three classes of problems: dense linear algebra, sparse direct methods and fast multipole methods. The resulting codes have been incorporated into Magma, Pastix and ScalFMM solvers, respectively. {\textcopyright} 2012 IEEE.",

author = "Emmanuel Agullo and George Bosilca and B{\'e}renger Bramas and Cedric Castagnede and Olivier Coulaud and Darve, {Eric F.} and Jack Dongarra and Mathieu Faverge and Nathalie Furmento and Luc Giraud and Xavier Lacoste and Julien Langou and Hatem Ltaief and Matthias Messner and Raymond Namyst and Pierre Ramet and Toru Takahashi and Samuel Thibault and Tomov, {Stanimire Z.} and Ichitaro Yamazaki",

note = "KAUST Repository Item: Exported on 2020-10-01",

year = "2012",

month = nov,

doi = "10.1109/SC.Companion.2012.167",

language = "English (US)",

isbn = "9780769549569",

pages = "1330--1332",

booktitle = "2012 SC Companion: High Performance Computing, Networking Storage and Analysis",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

}

Agullo, E, Bosilca, G, Bramas, B, Castagnede, C, Coulaud, O, Darve, EF, Dongarra, J, Faverge, M, Furmento, N, Giraud, L, Lacoste, X, Langou, J, Ltaief, H, Messner, M, Namyst, R, Ramet, P, Takahashi, T, Thibault, S, Tomov, SZ & Yamazaki, I 2012, Matrices over runtime systems at exascale. in 2012 SC Companion: High Performance Computing, Networking Storage and Analysis. Institute of Electrical and Electronics Engineers (IEEE), pp. 1330-1332. https://doi.org/10.1109/SC.Companion.2012.167

TY - GEN

T1 - Matrices over runtime systems at exascale

AU - Agullo, Emmanuel

AU - Bosilca, George

AU - Bramas, Bérenger

AU - Castagnede, Cedric

AU - Coulaud, Olivier

AU - Darve, Eric F.

AU - Dongarra, Jack

AU - Faverge, Mathieu

AU - Furmento, Nathalie

AU - Giraud, Luc

AU - Lacoste, Xavier

AU - Langou, Julien

AU - Ltaief, Hatem

AU - Messner, Matthias

AU - Namyst, Raymond

AU - Ramet, Pierre

AU - Takahashi, Toru

AU - Thibault, Samuel

AU - Tomov, Stanimire Z.

AU - Yamazaki, Ichitaro

N1 - KAUST Repository Item: Exported on 2020-10-01

PY - 2012/11

Y1 - 2012/11

N2 - The goal of Matrices Over Runtime Systems at Exascale (MORSE) project is to design dense and sparse linear algebra methods that achieve the fastest possible time to an accurate solution on large-scale multicore systems with GPU accelerators, using all the processing power that future high end systems can make available. In this poster, we propose a framework for describing linear algebra algorithms at a high level of abstraction and delegating the actual execution to a runtime system in order to design software whose performance is portable accross architectures. We illustrate our methodology on three classes of problems: dense linear algebra, sparse direct methods and fast multipole methods. The resulting codes have been incorporated into Magma, Pastix and ScalFMM solvers, respectively. © 2012 IEEE.

AB - The goal of Matrices Over Runtime Systems at Exascale (MORSE) project is to design dense and sparse linear algebra methods that achieve the fastest possible time to an accurate solution on large-scale multicore systems with GPU accelerators, using all the processing power that future high end systems can make available. In this poster, we propose a framework for describing linear algebra algorithms at a high level of abstraction and delegating the actual execution to a runtime system in order to design software whose performance is portable accross architectures. We illustrate our methodology on three classes of problems: dense linear algebra, sparse direct methods and fast multipole methods. The resulting codes have been incorporated into Magma, Pastix and ScalFMM solvers, respectively. © 2012 IEEE.

UR - http://hdl.handle.net/10754/575809

UR - http://ieeexplore.ieee.org/document/6495950/

UR - http://www.scopus.com/inward/record.url?scp=84876537122&partnerID=8YFLogxK

U2 - 10.1109/SC.Companion.2012.167

DO - 10.1109/SC.Companion.2012.167

M3 - Conference contribution

SN - 9780769549569

SP - 1330

EP - 1332

BT - 2012 SC Companion: High Performance Computing, Networking Storage and Analysis

PB - Institute of Electrical and Electronics Engineers (IEEE)

ER -

Matrices over runtime systems at exascale

Abstract

Access to Document

Other files and links

Fingerprint

Cite this