GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials

Markus Eisenbach; Jeff Larkin; Justin Lutjens; Steven Rennich; James H. Rogers

doi:10.1007/978-981-10-0457-5_24

GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials

Markus Eisenbach, Jeff Larkin, Justin Lutjens, Steven Rennich, James H. Rogers

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

Abstract

The Locally Self-consistent Multiple Scattering (LSMS) code solves the first principles Density Functional theory Kohn-Sham equation for a wide range of materials with a special focus on metals, alloys and metallic nano-structures. It has traditionally exhibited near perfect scalability on massively parallel high performance computer architectures. We present our efforts to exploit GPUs to accelerate the LSMS code to enable first principles calculations of O(100,000) atoms and statistical physics sampling of finite temperature properties. Using the Cray XK7 system Titan at the Oak Ridge Leadership Computing Facility we achieve a sustained performance of 14.5PFlop/s and a speedup of 8.6 compared to the CPU only code.

Original language	English
Title of host publication	Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings
Editors	Gansen Zhao, Zeguang Lu, Wenguang Chen, Guisheng Yin, Qilong Han, Weipeng Jing, Guanglu Sun
Publisher	Springer Verlag
Pages	259-268
Number of pages	10
ISBN (Print)	9789811004568
DOIs	https://doi.org/10.1007/978-981-10-0457-5_24
State	Published - 2016
Event	1st National Conference on Big Data Technology and Applications, BDTA 2015 - Harbin, China Duration: Dec 25 2015 → Dec 26 2015

Publication series

Name	Communications in Computer and Information Science
Volume	590
ISSN (Print)	1865-0929

Conference

Conference	1st National Conference on Big Data Technology and Applications, BDTA 2015
Country/Territory	China
City	Harbin
Period	12/25/15 → 12/26/15

Funding

This work has been sponsored by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Material Sciences and Engineering Division (basic theory and applications) and by the Office of Advanced Scientific Computing (software optimization and performance measurements). This research used resources of the Oak Ridge Leadership Computing Facility, which is supported by the Office of Science of the U.S. Department of Energy under contract no. DE-AC05-00OR22725 .

Funders	Funder number
U.S. Department of Energy
Office of Science
Basic Energy Sciences
Advanced Scientific Computing Research

Access to Document

10.1007/978-981-10-0457-5_24

Cite this

Eisenbach, M., Larkin, J., Lutjens, J., Rennich, S., & Rogers, J. H. (2016). GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials. In G. Zhao, Z. Lu, W. Chen, G. Yin, Q. Han, W. Jing, & G. Sun (Eds.), Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings (pp. 259-268). (Communications in Computer and Information Science; Vol. 590). Springer Verlag. https://doi.org/10.1007/978-981-10-0457-5_24

Eisenbach, Markus ; Larkin, Jeff ; Lutjens, Justin et al. / GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials. Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings. editor / Gansen Zhao ; Zeguang Lu ; Wenguang Chen ; Guisheng Yin ; Qilong Han ; Weipeng Jing ; Guanglu Sun. Springer Verlag, 2016. pp. 259-268 (Communications in Computer and Information Science).

@inproceedings{003bbaddb07d4754bb940658a0e601aa,

title = "GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials",

abstract = "The Locally Self-consistent Multiple Scattering (LSMS) code solves the first principles Density Functional theory Kohn-Sham equation for a wide range of materials with a special focus on metals, alloys and metallic nano-structures. It has traditionally exhibited near perfect scalability on massively parallel high performance computer architectures. We present our efforts to exploit GPUs to accelerate the LSMS code to enable first principles calculations of O(100,000) atoms and statistical physics sampling of finite temperature properties. Using the Cray XK7 system Titan at the Oak Ridge Leadership Computing Facility we achieve a sustained performance of 14.5PFlop/s and a speedup of 8.6 compared to the CPU only code.",

author = "Markus Eisenbach and Jeff Larkin and Justin Lutjens and Steven Rennich and Rogers, {James H.}",

note = "Publisher Copyright: {\textcopyright} Springer Science+Business Media Singapore 2016.; 1st National Conference on Big Data Technology and Applications, BDTA 2015 ; Conference date: 25-12-2015 Through 26-12-2015",

year = "2016",

doi = "10.1007/978-981-10-0457-5_24",

language = "English",

isbn = "9789811004568",

series = "Communications in Computer and Information Science",

publisher = "Springer Verlag",

pages = "259--268",

editor = "Gansen Zhao and Zeguang Lu and Wenguang Chen and Guisheng Yin and Qilong Han and Weipeng Jing and Guanglu Sun",

booktitle = "Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings",

}

Eisenbach, M, Larkin, J, Lutjens, J, Rennich, S & Rogers, JH 2016, GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials. in G Zhao, Z Lu, W Chen, G Yin, Q Han, W Jing & G Sun (eds), Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings. Communications in Computer and Information Science, vol. 590, Springer Verlag, pp. 259-268, 1st National Conference on Big Data Technology and Applications, BDTA 2015, Harbin, China, 12/25/15. https://doi.org/10.1007/978-981-10-0457-5_24

GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials. / Eisenbach, Markus; Larkin, Jeff; Lutjens, Justin et al.
Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings. ed. / Gansen Zhao; Zeguang Lu; Wenguang Chen; Guisheng Yin; Qilong Han; Weipeng Jing; Guanglu Sun. Springer Verlag, 2016. p. 259-268 (Communications in Computer and Information Science; Vol. 590).

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

TY - GEN

T1 - GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials

AU - Eisenbach, Markus

AU - Larkin, Jeff

AU - Lutjens, Justin

AU - Rennich, Steven

AU - Rogers, James H.

N1 - Publisher Copyright: © Springer Science+Business Media Singapore 2016.

PY - 2016

Y1 - 2016

N2 - The Locally Self-consistent Multiple Scattering (LSMS) code solves the first principles Density Functional theory Kohn-Sham equation for a wide range of materials with a special focus on metals, alloys and metallic nano-structures. It has traditionally exhibited near perfect scalability on massively parallel high performance computer architectures. We present our efforts to exploit GPUs to accelerate the LSMS code to enable first principles calculations of O(100,000) atoms and statistical physics sampling of finite temperature properties. Using the Cray XK7 system Titan at the Oak Ridge Leadership Computing Facility we achieve a sustained performance of 14.5PFlop/s and a speedup of 8.6 compared to the CPU only code.

AB - The Locally Self-consistent Multiple Scattering (LSMS) code solves the first principles Density Functional theory Kohn-Sham equation for a wide range of materials with a special focus on metals, alloys and metallic nano-structures. It has traditionally exhibited near perfect scalability on massively parallel high performance computer architectures. We present our efforts to exploit GPUs to accelerate the LSMS code to enable first principles calculations of O(100,000) atoms and statistical physics sampling of finite temperature properties. Using the Cray XK7 system Titan at the Oak Ridge Leadership Computing Facility we achieve a sustained performance of 14.5PFlop/s and a speedup of 8.6 compared to the CPU only code.

UR - http://www.scopus.com/inward/record.url?scp=84959018416&partnerID=8YFLogxK

U2 - 10.1007/978-981-10-0457-5_24

DO - 10.1007/978-981-10-0457-5_24

M3 - Conference contribution

AN - SCOPUS:84959018416

SN - 9789811004568

T3 - Communications in Computer and Information Science

SP - 259

EP - 268

BT - Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings

A2 - Zhao, Gansen

A2 - Lu, Zeguang

A2 - Chen, Wenguang

A2 - Yin, Guisheng

A2 - Han, Qilong

A2 - Jing, Weipeng

A2 - Sun, Guanglu

PB - Springer Verlag

T2 - 1st National Conference on Big Data Technology and Applications, BDTA 2015

Y2 - 25 December 2015 through 26 December 2015

ER -

Eisenbach M, Larkin J, Lutjens J, Rennich S, Rogers JH. GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials. In Zhao G, Lu Z, Chen W, Yin G, Han Q, Jing W, Sun G, editors, Big Data Technology and Applications - 1st National Conference, BDTA 2015, Proceedings. Springer Verlag. 2016. p. 259-268. (Communications in Computer and Information Science). doi: 10.1007/978-981-10-0457-5_24

GPU acceleration of the locally selfconsistent multiple scattering code for first principles calculation of the ground state and statistical physics of materials

Abstract

Publication series

Conference

Funding

Access to Document

Other files and links

Fingerprint

Cite this