Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions

Jonathan Eastep; Steve Sylvester; Christopher Cantalupo; Brad Geltz; Federico Ardanaz; Asma Al-Rawi; Kelly Livingston; Fuat Keceli; Matthias Maiterth; Siddhartha Jana

doi:10.1007/978-3-319-58667-0_21

Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions

Jonathan Eastep, Steve Sylvester, Christopher Cantalupo, Brad Geltz, Federico Ardanaz, Asma Al-Rawi, Kelly Livingston, Fuat Keceli, Matthias Maiterth, Siddhartha Jana

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

72 Scopus citations

Abstract

The power scaling challenge associated with Exascale systems is a well-known issue. In this work, we introduce the Global Extensible Open Power Manager (GEOPM): a tree-hierarchical, open source runtime framework we are contributing to the HPC community to foster increased collaboration and accelerated progress toward software-hardware co-designed energy management solutions that address Exascale power challenges and improve performance and energy efficiency in current systems. Through its plugin extensible architecture, GEOPM enables rapid prototyping of new energy management strategies. Different plugins can be tailored to the specific performance or energy efficiency priorities of each HPC center. To demonstrate the potential of the framework, this work develops an example plugin for GEOPM. This power rebalancing plugin targets power-capped systems and improves efficiency by minimizing job time-to-solution within a power budget. Our results demonstrate up to 30% improvements in the time-to-solution of CORAL system procurement benchmarks on a Xeon Phi cluster.

Original language	English
Title of host publication	High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings
Editors	Julian M. Kunkel, Pavan Balaji, David Keyes, Rio Yokota
Publisher	Springer Verlag
Pages	394-412
Number of pages	19
ISBN (Print)	9783319586663
DOIs	https://doi.org/10.1007/978-3-319-58667-0_21
State	Published - 2017
Externally published	Yes
Event	32nd International Conference, ISC High Performance, 2017 - Frankfurt, Germany Duration: Jun 18 2017 → Jun 22 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10266 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	32nd International Conference, ISC High Performance, 2017
Country/Territory	Germany
City	Frankfurt
Period	06/18/17 → 06/22/17

Funding

The authors would like to thank the following individuals for their input on this work: Vitali Morozov and Kalyan Kumaran of Argonne; Barry Rountree, Martin Schulz, and their teams from LLNL; James Laros, Ryan Grant, and their team from Sandia; and Richard Greco, Tryggve Fossum, David Lombard, Michael Patterson, and Alan Gara of Intel. Development of the GEOPM software package has been partially funded through contract B609815 with Argonne National Laboratory.

Access to Document

10.1007/978-3-319-58667-0_21

Cite this

Eastep, J., Sylvester, S., Cantalupo, C., Geltz, B., Ardanaz, F., Al-Rawi, A., Livingston, K., Keceli, F., Maiterth, M., & Jana, S. (2017). Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions. In J. M. Kunkel, P. Balaji, D. Keyes, & R. Yokota (Eds.), High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings (pp. 394-412). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10266 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-58667-0_21

Eastep, Jonathan ; Sylvester, Steve ; Cantalupo, Christopher et al. / Global extensible open power manager : A vehicle for HPC community collaboration on co-designed energy management solutions. High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings. editor / Julian M. Kunkel ; Pavan Balaji ; David Keyes ; Rio Yokota. Springer Verlag, 2017. pp. 394-412 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{5fe7e62da2804a97a4af0a3016321fc2,

title = "Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions",

abstract = "The power scaling challenge associated with Exascale systems is a well-known issue. In this work, we introduce the Global Extensible Open Power Manager (GEOPM): a tree-hierarchical, open source runtime framework we are contributing to the HPC community to foster increased collaboration and accelerated progress toward software-hardware co-designed energy management solutions that address Exascale power challenges and improve performance and energy efficiency in current systems. Through its plugin extensible architecture, GEOPM enables rapid prototyping of new energy management strategies. Different plugins can be tailored to the specific performance or energy efficiency priorities of each HPC center. To demonstrate the potential of the framework, this work develops an example plugin for GEOPM. This power rebalancing plugin targets power-capped systems and improves efficiency by minimizing job time-to-solution within a power budget. Our results demonstrate up to 30% improvements in the time-to-solution of CORAL system procurement benchmarks on a Xeon Phi cluster.",

author = "Jonathan Eastep and Steve Sylvester and Christopher Cantalupo and Brad Geltz and Federico Ardanaz and Asma Al-Rawi and Kelly Livingston and Fuat Keceli and Matthias Maiterth and Siddhartha Jana",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 32nd International Conference, ISC High Performance, 2017 ; Conference date: 18-06-2017 Through 22-06-2017",

year = "2017",

doi = "10.1007/978-3-319-58667-0_21",

language = "English",

isbn = "9783319586663",

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

publisher = "Springer Verlag",

pages = "394--412",

editor = "Kunkel, {Julian M.} and Pavan Balaji and David Keyes and Rio Yokota",

booktitle = "High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings",

}

Eastep, J, Sylvester, S, Cantalupo, C, Geltz, B, Ardanaz, F, Al-Rawi, A, Livingston, K, Keceli, F, Maiterth, M & Jana, S 2017, Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions. in JM Kunkel, P Balaji, D Keyes & R Yokota (eds), High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10266 LNCS, Springer Verlag, pp. 394-412, 32nd International Conference, ISC High Performance, 2017, Frankfurt, Germany, 06/18/17. https://doi.org/10.1007/978-3-319-58667-0_21

Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions. / Eastep, Jonathan; Sylvester, Steve; Cantalupo, Christopher et al.
High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings. ed. / Julian M. Kunkel; Pavan Balaji; David Keyes; Rio Yokota. Springer Verlag, 2017. p. 394-412 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10266 LNCS).

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

TY - GEN

T1 - Global extensible open power manager

T2 - 32nd International Conference, ISC High Performance, 2017

AU - Eastep, Jonathan

AU - Sylvester, Steve

AU - Cantalupo, Christopher

AU - Geltz, Brad

AU - Ardanaz, Federico

AU - Al-Rawi, Asma

AU - Livingston, Kelly

AU - Keceli, Fuat

AU - Maiterth, Matthias

AU - Jana, Siddhartha

PY - 2017

Y1 - 2017

N2 - The power scaling challenge associated with Exascale systems is a well-known issue. In this work, we introduce the Global Extensible Open Power Manager (GEOPM): a tree-hierarchical, open source runtime framework we are contributing to the HPC community to foster increased collaboration and accelerated progress toward software-hardware co-designed energy management solutions that address Exascale power challenges and improve performance and energy efficiency in current systems. Through its plugin extensible architecture, GEOPM enables rapid prototyping of new energy management strategies. Different plugins can be tailored to the specific performance or energy efficiency priorities of each HPC center. To demonstrate the potential of the framework, this work develops an example plugin for GEOPM. This power rebalancing plugin targets power-capped systems and improves efficiency by minimizing job time-to-solution within a power budget. Our results demonstrate up to 30% improvements in the time-to-solution of CORAL system procurement benchmarks on a Xeon Phi cluster.

AB - The power scaling challenge associated with Exascale systems is a well-known issue. In this work, we introduce the Global Extensible Open Power Manager (GEOPM): a tree-hierarchical, open source runtime framework we are contributing to the HPC community to foster increased collaboration and accelerated progress toward software-hardware co-designed energy management solutions that address Exascale power challenges and improve performance and energy efficiency in current systems. Through its plugin extensible architecture, GEOPM enables rapid prototyping of new energy management strategies. Different plugins can be tailored to the specific performance or energy efficiency priorities of each HPC center. To demonstrate the potential of the framework, this work develops an example plugin for GEOPM. This power rebalancing plugin targets power-capped systems and improves efficiency by minimizing job time-to-solution within a power budget. Our results demonstrate up to 30% improvements in the time-to-solution of CORAL system procurement benchmarks on a Xeon Phi cluster.

UR - http://www.scopus.com/inward/record.url?scp=85026247920&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-58667-0_21

DO - 10.1007/978-3-319-58667-0_21

M3 - Conference contribution

AN - SCOPUS:85026247920

SN - 9783319586663

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

SP - 394

EP - 412

BT - High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings

A2 - Kunkel, Julian M.

A2 - Balaji, Pavan

A2 - Keyes, David

A2 - Yokota, Rio

PB - Springer Verlag

Y2 - 18 June 2017 through 22 June 2017

ER -

Eastep J, Sylvester S, Cantalupo C, Geltz B, Ardanaz F, Al-Rawi A et al. Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions. In Kunkel JM, Balaji P, Keyes D, Yokota R, editors, High Performance Computing - 32nd International Conference, ISC High Performance 2017, Proceedings. Springer Verlag. 2017. p. 394-412. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-58667-0_21

Global extensible open power manager: A vehicle for HPC community collaboration on co-designed energy management solutions

Abstract

Publication series

Conference

Funding

Access to Document

Other files and links

Fingerprint

Cite this