skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Early Experiences with Node-Level Power Capping on the Cray XC40 Platform

Abstract

Power consumption of extreme-scale supercomputers has become a key performance bottleneck. Yet current practices do not leverage power management opportunities, instead running at ''maximum power''. This is not sustainable. Future systems will need to manage power as a critical resource, directing where it has greatest benefit. Power capping is one mechanism for managing power budgets, however its behavior is not well understood. This paper presents an empirical evaluation of several key HPC workloads running under a power cap on a Cray XC40 system, and provides a comparison of this technique with p-state control, demonstrating the performance differences of each. These results show: 1. Maximum performance requires ensuring the cap is not reached; 2. Performance slowdown under a cap can be attributed to cascading delays which result in unsynchronized performance variability across nodes; and, 3. Due to lag in reaction time, considerable time is spent operating above the set cap. This work provides a timely and much needed comparison of HPC application performance under a power cap and attempts to enable users and system administrators to understand how to best optimize application performance on power-constrained HPC systems.

Authors:
; ; ; ;
Publication Date:
Research Org.:
Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
Sponsoring Org.:
USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1338038
Report Number(s):
SAND2015-8821C
614669
DOE Contract Number:  
AC04-94AL85000
Resource Type:
Conference
Resource Relation:
Conference: Proposed for presentation at the Workshop on Energy Efficient Supercomputing (E2SC) held November 15 - October 15, 2015 in Austin, TX.
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING

Citation Formats

Pedretti, Kevin, Olivier, Stephen Lecler, Ferreira, Kurt Brian, Shipman, Galen, and Shu, Wei. Early Experiences with Node-Level Power Capping on the Cray XC40 Platform. United States: N. p., 2015. Web. doi:10.1145/2834800.2834801.
Pedretti, Kevin, Olivier, Stephen Lecler, Ferreira, Kurt Brian, Shipman, Galen, & Shu, Wei. Early Experiences with Node-Level Power Capping on the Cray XC40 Platform. United States. doi:10.1145/2834800.2834801.
Pedretti, Kevin, Olivier, Stephen Lecler, Ferreira, Kurt Brian, Shipman, Galen, and Shu, Wei. Thu . "Early Experiences with Node-Level Power Capping on the Cray XC40 Platform". United States. doi:10.1145/2834800.2834801. https://www.osti.gov/servlets/purl/1338038.
@article{osti_1338038,
title = {Early Experiences with Node-Level Power Capping on the Cray XC40 Platform},
author = {Pedretti, Kevin and Olivier, Stephen Lecler and Ferreira, Kurt Brian and Shipman, Galen and Shu, Wei},
abstractNote = {Power consumption of extreme-scale supercomputers has become a key performance bottleneck. Yet current practices do not leverage power management opportunities, instead running at ''maximum power''. This is not sustainable. Future systems will need to manage power as a critical resource, directing where it has greatest benefit. Power capping is one mechanism for managing power budgets, however its behavior is not well understood. This paper presents an empirical evaluation of several key HPC workloads running under a power cap on a Cray XC40 system, and provides a comparison of this technique with p-state control, demonstrating the performance differences of each. These results show: 1. Maximum performance requires ensuring the cap is not reached; 2. Performance slowdown under a cap can be attributed to cascading delays which result in unsynchronized performance variability across nodes; and, 3. Due to lag in reaction time, considerable time is spent operating above the set cap. This work provides a timely and much needed comparison of HPC application performance under a power cap and attempts to enable users and system administrators to understand how to best optimize application performance on power-constrained HPC systems.},
doi = {10.1145/2834800.2834801},
journal = {},
number = ,
volume = ,
place = {United States},
year = {2015},
month = {10}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share: