Parallel electronic structure calculations using multiple graphics processing units (GPUs)

Samuli Hakala; Ville Havu; Jussi Enkovaara; Risto Nieminen

doi:10.1007/978-3-642-36803-5_4

Parallel electronic structure calculations using multiple graphics processing units (GPUs)

Samuli Hakala^*, Ville Havu, Jussi Enkovaara, Risto Nieminen

^*Corresponding author for this work

Department of Applied Physics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

17 Citations (Scopus)

Abstract

We present an implementation of parallel GPU-accelerated GPAW, a density-functional theory (DFT) code based on grid based projector-augmented wave method. GPAW is suitable for large scale electronic structure calculations and capable of scaling to thousands of cores. We have accelerated the most computationally intensive components of the program with CUDA. We will provide performance and scaling analysis of our multi-GPU-accelerated code staring from small systems up to systems with thousands of atoms running on GPU clusters. We have achieved up to 15 times speed-ups on large systems.

Original language	English
Title of host publication	Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers
Pages	63-76
Number of pages	14
Volume	7782 LNCS
DOIs	https://doi.org/10.1007/978-3-642-36803-5_4
Publication status	Published - 2013
MoE publication type	A4 Article in a conference publication
Event	International Conference on Applied Parallel and Scientific Computing - Helsinki, Finland Duration: 10 Jun 2012 → 13 Jun 2012 Conference number: 11

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7782 LNCS
ISSN (Print)	03029743
ISSN (Electronic)	16113349

Conference

Conference	International Conference on Applied Parallel and Scientific Computing
Abbreviated title	PARA
Country/Territory	Finland
City	Helsinki
Period	10/06/2012 → 13/06/2012

Keywords

density functional theory
electronic structure calculations
graphics processing units

Access to Document

10.1007/978-3-642-36803-5_4

OpenUrl availability

Full text

Cite this

Hakala, S., Havu, V., Enkovaara, J., & Nieminen, R. (2013). Parallel electronic structure calculations using multiple graphics processing units (GPUs). In Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers (Vol. 7782 LNCS, pp. 63-76). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7782 LNCS). https://doi.org/10.1007/978-3-642-36803-5_4

Hakala, Samuli ; Havu, Ville ; Enkovaara, Jussi et al. / Parallel electronic structure calculations using multiple graphics processing units (GPUs). Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers. Vol. 7782 LNCS 2013. pp. 63-76 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{5923deacaf674477946be93da50b5ce3,

title = "Parallel electronic structure calculations using multiple graphics processing units (GPUs)",

abstract = "We present an implementation of parallel GPU-accelerated GPAW, a density-functional theory (DFT) code based on grid based projector-augmented wave method. GPAW is suitable for large scale electronic structure calculations and capable of scaling to thousands of cores. We have accelerated the most computationally intensive components of the program with CUDA. We will provide performance and scaling analysis of our multi-GPU-accelerated code staring from small systems up to systems with thousands of atoms running on GPU clusters. We have achieved up to 15 times speed-ups on large systems.",

keywords = "density functional theory, electronic structure calculations, graphics processing units",

author = "Samuli Hakala and Ville Havu and Jussi Enkovaara and Risto Nieminen",

year = "2013",

doi = "10.1007/978-3-642-36803-5_4",

language = "English",

isbn = "9783642368028",

volume = "7782 LNCS",

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

pages = "63--76",

booktitle = "Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers",

note = "International Conference on Applied Parallel and Scientific Computing, PARA ; Conference date: 10-06-2012 Through 13-06-2012",

}

Hakala, S , Havu, V, Enkovaara, J & Nieminen, R 2013, Parallel electronic structure calculations using multiple graphics processing units (GPUs). in Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers. vol. 7782 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7782 LNCS, pp. 63-76, International Conference on Applied Parallel and Scientific Computing, Helsinki, Finland, 10/06/2012. https://doi.org/10.1007/978-3-642-36803-5_4

Parallel electronic structure calculations using multiple graphics processing units (GPUs). / Hakala, Samuli ; Havu, Ville; Enkovaara, Jussi et al.

Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers. Vol. 7782 LNCS 2013. p. 63-76 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7782 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Parallel electronic structure calculations using multiple graphics processing units (GPUs)

AU - Hakala, Samuli

AU - Havu, Ville

AU - Enkovaara, Jussi

AU - Nieminen, Risto

N1 - Conference code: 11

PY - 2013

Y1 - 2013

N2 - We present an implementation of parallel GPU-accelerated GPAW, a density-functional theory (DFT) code based on grid based projector-augmented wave method. GPAW is suitable for large scale electronic structure calculations and capable of scaling to thousands of cores. We have accelerated the most computationally intensive components of the program with CUDA. We will provide performance and scaling analysis of our multi-GPU-accelerated code staring from small systems up to systems with thousands of atoms running on GPU clusters. We have achieved up to 15 times speed-ups on large systems.

AB - We present an implementation of parallel GPU-accelerated GPAW, a density-functional theory (DFT) code based on grid based projector-augmented wave method. GPAW is suitable for large scale electronic structure calculations and capable of scaling to thousands of cores. We have accelerated the most computationally intensive components of the program with CUDA. We will provide performance and scaling analysis of our multi-GPU-accelerated code staring from small systems up to systems with thousands of atoms running on GPU clusters. We have achieved up to 15 times speed-ups on large systems.

KW - density functional theory

KW - electronic structure calculations

KW - graphics processing units

UR - http://www.scopus.com/inward/record.url?scp=84874214919&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-36803-5_4

DO - 10.1007/978-3-642-36803-5_4

M3 - Conference contribution

AN - SCOPUS:84874214919

SN - 9783642368028

VL - 7782 LNCS

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

SP - 63

EP - 76

BT - Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers

T2 - International Conference on Applied Parallel and Scientific Computing

Y2 - 10 June 2012 through 13 June 2012

ER -

Hakala S , Havu V, Enkovaara J, Nieminen R. Parallel electronic structure calculations using multiple graphics processing units (GPUs). In Applied Parallel and Scientific Computing - 11th International Conference, PARA 2012, Revised Selected Papers. Vol. 7782 LNCS. 2013. p. 63-76. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-36803-5_4

Parallel electronic structure calculations using multiple graphics processing units (GPUs)

Abstract

Publication series

Conference

Keywords

Access to Document

OpenUrl availability

Other files and links

Fingerprint

Cite this