Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects

Andreas Alvermann; Achim Basermann; Hans-Joachim Bungartz; Christian Carbogno; Dominik Ernst; Holger Fehske; Yasunori Futamura; Martin Galgon; Georg Hager; Sarah Huber; Thomas Huckle; Akihiro Ida; Akira Imakura; Masatoshi Kawai; Simone Koecher; Moritz Kreutzer; Pavel Kus; Bruno Lang; Hermann Lederer; Valeriy Manin; Andreas Marek; Kengo Nakajima; Lydia Nemec; Karsten Reuter; Michael Rippl; Melven Roehrig-Zoellner; Tetsuya Sakurai; Matthias Scheffler; Christoph Scheurer; Faisal Shahzad; Danilo Simoes Brambila; Jonas Thies; Gerhard Wellein

doi:10.1007/s13160-019-00360-8

Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects

Andreas Alvermann, Achim Basermann, Hans-Joachim Bungartz, Christian Carbogno, Dominik Ernst, Holger Fehske, Yasunori Futamura, Martin Galgon, Georg Hager, Sarah Huber, Thomas Huckle, Akihiro Ida, Akira Imakura, Masatoshi Kawai, Simone Koecher, Moritz Kreutzer, Pavel Kus, Bruno Lang, Hermann Lederer, Valeriy ManinAndreas Marek, Kengo Nakajima, Lydia Nemec, Karsten Reuter, Michael Rippl, Melven Roehrig-Zoellner, Tetsuya Sakurai, Matthias Scheffler, Christoph Scheurer, Faisal Shahzad, Danilo Simoes Brambila, Jonas Thies, Gerhard Wellein

Research output: Contribution to journal › Article › Scientific › peer-review

9 Citations (Scopus)

Abstract

We first briefly report on the status and recent achievements of the ELPA-AEO (Eigen value Solvers for Petaflop Applications—Algorithmic Extensions and Optimizations) and ESSEX II (Equipping Sparse Solvers for Exascale) projects. In both collaboratory efforts, scientists from the application areas, mathematicians, and computer scientists work together to develop and make available efficient highly parallel methods for the solution of eigenvalue problems. Then we focus on a topic addressed in both projects, the use of mixed precision computations to enhance efficiency. We give a more detailed description of our approaches for benefiting from either lower or higher precision in three selected contexts and of the results thus obtained.

Original language	English
Pages (from-to)	699-717
Number of pages	19
Journal	JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS
Volume	36
Issue number	2
DOIs	https://doi.org/10.1007/s13160-019-00360-8
Publication status	Published - Jul 2019
Externally published	Yes

Keywords

ELPA-AEO
ESSEX
Eigensolver
Mixed precision
Parallel

Access to Document

10.1007/s13160-019-00360-8

Cite this

Alvermann, A., Basermann, A., Bungartz, H.-J., Carbogno, C., Ernst, D., Fehske, H., Futamura, Y., Galgon, M., Hager, G., Huber, S., Huckle, T., Ida, A., Imakura, A., Kawai, M., Koecher, S., Kreutzer, M., Kus, P., Lang, B., Lederer, H., ... Wellein, G. (2019). Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects. JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS, 36(2), 699-717. https://doi.org/10.1007/s13160-019-00360-8

@article{279e88c3648d4eb18019483e6abb4c2d,

title = "Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects",

abstract = "We first briefly report on the status and recent achievements of the ELPA-AEO (Eigen value Solvers for Petaflop Applications—Algorithmic Extensions and Optimizations) and ESSEX II (Equipping Sparse Solvers for Exascale) projects. In both collaboratory efforts, scientists from the application areas, mathematicians, and computer scientists work together to develop and make available efficient highly parallel methods for the solution of eigenvalue problems. Then we focus on a topic addressed in both projects, the use of mixed precision computations to enhance efficiency. We give a more detailed description of our approaches for benefiting from either lower or higher precision in three selected contexts and of the results thus obtained.",

keywords = "ELPA-AEO, ESSEX, Eigensolver, Mixed precision, Parallel",

author = "Andreas Alvermann and Achim Basermann and Hans-Joachim Bungartz and Christian Carbogno and Dominik Ernst and Holger Fehske and Yasunori Futamura and Martin Galgon and Georg Hager and Sarah Huber and Thomas Huckle and Akihiro Ida and Akira Imakura and Masatoshi Kawai and Simone Koecher and Moritz Kreutzer and Pavel Kus and Bruno Lang and Hermann Lederer and Valeriy Manin and Andreas Marek and Kengo Nakajima and Lydia Nemec and Karsten Reuter and Michael Rippl and Melven Roehrig-Zoellner and Tetsuya Sakurai and Matthias Scheffler and Christoph Scheurer and Faisal Shahzad and Brambila, {Danilo Simoes} and Jonas Thies and Gerhard Wellein",

year = "2019",

month = jul,

doi = "10.1007/s13160-019-00360-8",

language = "English",

volume = "36",

pages = "699--717",

journal = "JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS",

issn = "0916-7005",

publisher = "Springer",

number = "2",

}

Alvermann, A, Basermann, A, Bungartz, H-J, Carbogno, C, Ernst, D, Fehske, H, Futamura, Y, Galgon, M, Hager, G, Huber, S, Huckle, T, Ida, A, Imakura, A, Kawai, M, Koecher, S, Kreutzer, M, Kus, P, Lang, B, Lederer, H, Manin, V, Marek, A, Nakajima, K, Nemec, L, Reuter, K, Rippl, M, Roehrig-Zoellner, M, Sakurai, T, Scheffler, M, Scheurer, C, Shahzad, F, Brambila, DS, Thies, J & Wellein, G 2019, 'Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects', JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS, vol. 36, no. 2, pp. 699-717. https://doi.org/10.1007/s13160-019-00360-8

TY - JOUR

T1 - Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects

AU - Alvermann, Andreas

AU - Basermann, Achim

AU - Bungartz, Hans-Joachim

AU - Carbogno, Christian

AU - Ernst, Dominik

AU - Fehske, Holger

AU - Futamura, Yasunori

AU - Galgon, Martin

AU - Hager, Georg

AU - Huber, Sarah

AU - Huckle, Thomas

AU - Ida, Akihiro

AU - Imakura, Akira

AU - Kawai, Masatoshi

AU - Koecher, Simone

AU - Kreutzer, Moritz

AU - Kus, Pavel

AU - Lang, Bruno

AU - Lederer, Hermann

AU - Manin, Valeriy

AU - Marek, Andreas

AU - Nakajima, Kengo

AU - Nemec, Lydia

AU - Reuter, Karsten

AU - Rippl, Michael

AU - Roehrig-Zoellner, Melven

AU - Sakurai, Tetsuya

AU - Scheffler, Matthias

AU - Scheurer, Christoph

AU - Shahzad, Faisal

AU - Brambila, Danilo Simoes

AU - Thies, Jonas

AU - Wellein, Gerhard

PY - 2019/7

Y1 - 2019/7

N2 - We first briefly report on the status and recent achievements of the ELPA-AEO (Eigen value Solvers for Petaflop Applications—Algorithmic Extensions and Optimizations) and ESSEX II (Equipping Sparse Solvers for Exascale) projects. In both collaboratory efforts, scientists from the application areas, mathematicians, and computer scientists work together to develop and make available efficient highly parallel methods for the solution of eigenvalue problems. Then we focus on a topic addressed in both projects, the use of mixed precision computations to enhance efficiency. We give a more detailed description of our approaches for benefiting from either lower or higher precision in three selected contexts and of the results thus obtained.

AB - We first briefly report on the status and recent achievements of the ELPA-AEO (Eigen value Solvers for Petaflop Applications—Algorithmic Extensions and Optimizations) and ESSEX II (Equipping Sparse Solvers for Exascale) projects. In both collaboratory efforts, scientists from the application areas, mathematicians, and computer scientists work together to develop and make available efficient highly parallel methods for the solution of eigenvalue problems. Then we focus on a topic addressed in both projects, the use of mixed precision computations to enhance efficiency. We give a more detailed description of our approaches for benefiting from either lower or higher precision in three selected contexts and of the results thus obtained.

KW - ELPA-AEO

KW - ESSEX

KW - Eigensolver

KW - Mixed precision

KW - Parallel

UR - http://www.scopus.com/inward/record.url?scp=85064956441&partnerID=8YFLogxK

U2 - 10.1007/s13160-019-00360-8

DO - 10.1007/s13160-019-00360-8

M3 - Article

SN - 0916-7005

VL - 36

SP - 699

EP - 717

JO - JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS

JF - JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS

IS - 2

ER -

Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this