Essex: Equipping sparse solvers for exascale

Andreas Alvermann; Achim Basermann; Holger Fehske; Andreas Pieper; Melven Röhrig-Zöllner; Jonas Thies; Martin Galgon; Lukas Krämer; Bruno Lang; Georg Hager; Moritz Kreutzer; Faisal Shahzad; Gerhard Wellein

doi:10.1007/978-3-319-14313-2_49

Essex: Equipping sparse solvers for exascale

Andreas Alvermann^*, Achim Basermann, Holger Fehske, Andreas Pieper, Melven Röhrig-Zöllner, Jonas Thies, Martin Galgon, Lukas Krämer, Bruno Lang, Georg Hager, Moritz Kreutzer, Faisal Shahzad, Gerhard Wellein

^*Corresponding author for this work

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

5 Citations (Scopus)

Abstract

The ESSEX project investigates computational issues arising at exascale for large-scale sparse eigenvalue problems and develops programming concepts and numerical methods for their solution. The project pursues a coherent co-design of all software layers where a holistic performance engineering process guides code development across the classic boundaries of application, numerical method, and basic kernel library. Within ESSEX the numerical methods cover widely applicable solvers such as classic Krylov, Jacobi-Davidson, or the recent FEAST methods, as well as domain-specific iterative schemes relevant for the ESSEX quantum physics application. This report introduces the project structure and presents selected results which demonstrate the potential impact of ESSEX for efficient sparse solvers on highly scalable heterogeneous supercomputers.

Original language	English
Title of host publication	Euro-Par 2014
Subtitle of host publication	Parallel Processing Workshops - Euro-Par 2014 InternationalWorkshops, Revised Selected Papers
Editors	Luís Lopes
Publisher	Springer
Pages	577-588
Number of pages	12
ISBN (Electronic)	9783319143125
DOIs	https://doi.org/10.1007/978-3-319-14313-2_49
Publication status	Published - 2014
Externally published	Yes
Event	International Workshop on Parallel Processing, Euro-Par 2014 - Porto, Portugal Duration: 25 Aug 2014 → 26 Aug 2014

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	8806
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Workshop on Parallel Processing, Euro-Par 2014
Country/Territory	Portugal
City	Porto
Period	25/08/14 → 26/08/14

Access to Document

10.1007/978-3-319-14313-2_49

Cite this

Alvermann, A., Basermann, A., Fehske, H., Pieper, A., Röhrig-Zöllner, M., Thies, J., Galgon, M., Krämer, L., Lang, B., Hager, G., Kreutzer, M., Shahzad, F., & Wellein, G. (2014). Essex: Equipping sparse solvers for exascale. In L. Lopes (Ed.), Euro-Par 2014: Parallel Processing Workshops - Euro-Par 2014 InternationalWorkshops, Revised Selected Papers (pp. 577-588). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8806). Springer. https://doi.org/10.1007/978-3-319-14313-2_49

Alvermann, Andreas ; Basermann, Achim ; Fehske, Holger et al. / Essex : Equipping sparse solvers for exascale. Euro-Par 2014: Parallel Processing Workshops - Euro-Par 2014 InternationalWorkshops, Revised Selected Papers. editor / Luís Lopes. Springer, 2014. pp. 577-588 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{77aae50c98b64dfa84b0fe23278ded11,

title = "Essex: Equipping sparse solvers for exascale",

abstract = "The ESSEX project investigates computational issues arising at exascale for large-scale sparse eigenvalue problems and develops programming concepts and numerical methods for their solution. The project pursues a coherent co-design of all software layers where a holistic performance engineering process guides code development across the classic boundaries of application, numerical method, and basic kernel library. Within ESSEX the numerical methods cover widely applicable solvers such as classic Krylov, Jacobi-Davidson, or the recent FEAST methods, as well as domain-specific iterative schemes relevant for the ESSEX quantum physics application. This report introduces the project structure and presents selected results which demonstrate the potential impact of ESSEX for efficient sparse solvers on highly scalable heterogeneous supercomputers.",

author = "Andreas Alvermann and Achim Basermann and Holger Fehske and Andreas Pieper and Melven R{\"o}hrig-Z{\"o}llner and Jonas Thies and Martin Galgon and Lukas Kr{\"a}mer and Bruno Lang and Georg Hager and Moritz Kreutzer and Faisal Shahzad and Gerhard Wellein",

year = "2014",

doi = "10.1007/978-3-319-14313-2_49",

language = "English",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "577--588",

editor = "Lu{\'i}s Lopes",

booktitle = "Euro-Par 2014",

note = "International Workshop on Parallel Processing, Euro-Par 2014 ; Conference date: 25-08-2014 Through 26-08-2014",

}

Alvermann, A, Basermann, A, Fehske, H, Pieper, A, Röhrig-Zöllner, M, Thies, J, Galgon, M, Krämer, L, Lang, B, Hager, G, Kreutzer, M, Shahzad, F & Wellein, G 2014, Essex: Equipping sparse solvers for exascale. in L Lopes (ed.), Euro-Par 2014: Parallel Processing Workshops - Euro-Par 2014 InternationalWorkshops, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8806, Springer, pp. 577-588, International Workshop on Parallel Processing, Euro-Par 2014, Porto, Portugal, 25/08/14. https://doi.org/10.1007/978-3-319-14313-2_49

Essex: Equipping sparse solvers for exascale. / Alvermann, Andreas; Basermann, Achim; Fehske, Holger et al.
Euro-Par 2014: Parallel Processing Workshops - Euro-Par 2014 InternationalWorkshops, Revised Selected Papers. ed. / Luís Lopes. Springer, 2014. p. 577-588 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8806).

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Essex

T2 - International Workshop on Parallel Processing, Euro-Par 2014

AU - Alvermann, Andreas

AU - Basermann, Achim

AU - Fehske, Holger

AU - Pieper, Andreas

AU - Röhrig-Zöllner, Melven

AU - Thies, Jonas

AU - Galgon, Martin

AU - Krämer, Lukas

AU - Lang, Bruno

AU - Hager, Georg

AU - Kreutzer, Moritz

AU - Shahzad, Faisal

AU - Wellein, Gerhard

PY - 2014

Y1 - 2014

N2 - The ESSEX project investigates computational issues arising at exascale for large-scale sparse eigenvalue problems and develops programming concepts and numerical methods for their solution. The project pursues a coherent co-design of all software layers where a holistic performance engineering process guides code development across the classic boundaries of application, numerical method, and basic kernel library. Within ESSEX the numerical methods cover widely applicable solvers such as classic Krylov, Jacobi-Davidson, or the recent FEAST methods, as well as domain-specific iterative schemes relevant for the ESSEX quantum physics application. This report introduces the project structure and presents selected results which demonstrate the potential impact of ESSEX for efficient sparse solvers on highly scalable heterogeneous supercomputers.

AB - The ESSEX project investigates computational issues arising at exascale for large-scale sparse eigenvalue problems and develops programming concepts and numerical methods for their solution. The project pursues a coherent co-design of all software layers where a holistic performance engineering process guides code development across the classic boundaries of application, numerical method, and basic kernel library. Within ESSEX the numerical methods cover widely applicable solvers such as classic Krylov, Jacobi-Davidson, or the recent FEAST methods, as well as domain-specific iterative schemes relevant for the ESSEX quantum physics application. This report introduces the project structure and presents selected results which demonstrate the potential impact of ESSEX for efficient sparse solvers on highly scalable heterogeneous supercomputers.

UR - http://www.scopus.com/inward/record.url?scp=84920102873&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-14313-2_49

DO - 10.1007/978-3-319-14313-2_49

M3 - Conference contribution

AN - SCOPUS:84920102873

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 577

EP - 588

BT - Euro-Par 2014

A2 - Lopes, Luís

PB - Springer

Y2 - 25 August 2014 through 26 August 2014

ER -

Alvermann A, Basermann A, Fehske H, Pieper A, Röhrig-Zöllner M, Thies J et al. Essex: Equipping sparse solvers for exascale. In Lopes L, editor, Euro-Par 2014: Parallel Processing Workshops - Euro-Par 2014 InternationalWorkshops, Revised Selected Papers. Springer. 2014. p. 577-588. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-14313-2_49

Essex: Equipping sparse solvers for exascale

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this