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

Title: Spatiotemporal modeling of node temperatures in supercomputers

Abstract

Los Alamos National Laboratory (LANL) is home to many large supercomputing clusters. These clusters require an enormous amount of power (~500-2000 kW each), and most of this energy is converted into heat. Thus, cooling the components of the supercomputer becomes a critical and expensive endeavor. Recently a project was initiated to investigate the effect that changes to the cooling system in a machine room had on three large machines that were housed there. Coupled with this goal was the aim to develop a general good-practice for characterizing the effect of cooling changes and monitoring machine node temperatures in this and other machine rooms. This paper focuses on the statistical approach used to quantify the effect that several cooling changes to the room had on the temperatures of the individual nodes of the computers. The largest cluster in the room has 1,600 nodes that run a variety of jobs during general use. Since extremes temperatures are important, a Normal distribution plus generalized Pareto distribution for the upper tail is used to model the marginal distribution, along with a Gaussian process copula to account for spatio-temporal dependence. A Gaussian Markov random field (GMRF) model is used to model the spatial effects onmore » the node temperatures as the cooling changes take place. This model is then used to assess the condition of the node temperatures after each change to the room. The analysis approach was used to uncover the cause of a problematic episode of overheating nodes on one of the supercomputing clusters. Lastly, this same approach can easily be applied to monitor and investigate cooling systems at other data centers, as well.« less

Authors:
 [1];  [2];  [1];  [1];  [1];  [1];  [1]
  1. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
  2. North Carolina State Univ., Raleigh, NC (United States)
Publication Date:
Research Org.:
Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1329590
Report Number(s):
LA-UR-15-22229
Journal ID: ISSN 0162-1459
Grant/Contract Number:  
AC52-06NA25396
Resource Type:
Accepted Manuscript
Journal Name:
Journal of the American Statistical Association
Additional Journal Information:
Journal Name: Journal of the American Statistical Association; Journal ID: ISSN 0162-1459
Publisher:
Taylor & Francis
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING; Mathematics

Citation Formats

Storlie, Curtis Byron, Reich, Brian James, Rust, William Newton, Ticknor, Lawrence O., Bonnie, Amanda Marie, Montoya, Andrew J., and Michalak, Sarah E. Spatiotemporal modeling of node temperatures in supercomputers. United States: N. p., 2016. Web. doi:10.1080/01621459.2016.1195271.
Storlie, Curtis Byron, Reich, Brian James, Rust, William Newton, Ticknor, Lawrence O., Bonnie, Amanda Marie, Montoya, Andrew J., & Michalak, Sarah E. Spatiotemporal modeling of node temperatures in supercomputers. United States. doi:10.1080/01621459.2016.1195271.
Storlie, Curtis Byron, Reich, Brian James, Rust, William Newton, Ticknor, Lawrence O., Bonnie, Amanda Marie, Montoya, Andrew J., and Michalak, Sarah E. Fri . "Spatiotemporal modeling of node temperatures in supercomputers". United States. doi:10.1080/01621459.2016.1195271. https://www.osti.gov/servlets/purl/1329590.
@article{osti_1329590,
title = {Spatiotemporal modeling of node temperatures in supercomputers},
author = {Storlie, Curtis Byron and Reich, Brian James and Rust, William Newton and Ticknor, Lawrence O. and Bonnie, Amanda Marie and Montoya, Andrew J. and Michalak, Sarah E.},
abstractNote = {Los Alamos National Laboratory (LANL) is home to many large supercomputing clusters. These clusters require an enormous amount of power (~500-2000 kW each), and most of this energy is converted into heat. Thus, cooling the components of the supercomputer becomes a critical and expensive endeavor. Recently a project was initiated to investigate the effect that changes to the cooling system in a machine room had on three large machines that were housed there. Coupled with this goal was the aim to develop a general good-practice for characterizing the effect of cooling changes and monitoring machine node temperatures in this and other machine rooms. This paper focuses on the statistical approach used to quantify the effect that several cooling changes to the room had on the temperatures of the individual nodes of the computers. The largest cluster in the room has 1,600 nodes that run a variety of jobs during general use. Since extremes temperatures are important, a Normal distribution plus generalized Pareto distribution for the upper tail is used to model the marginal distribution, along with a Gaussian process copula to account for spatio-temporal dependence. A Gaussian Markov random field (GMRF) model is used to model the spatial effects on the node temperatures as the cooling changes take place. This model is then used to assess the condition of the node temperatures after each change to the room. The analysis approach was used to uncover the cause of a problematic episode of overheating nodes on one of the supercomputing clusters. Lastly, this same approach can easily be applied to monitor and investigate cooling systems at other data centers, as well.},
doi = {10.1080/01621459.2016.1195271},
journal = {Journal of the American Statistical Association},
number = ,
volume = ,
place = {United States},
year = {2016},
month = {6}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share:

Works referenced in this record:

A close look at the spatial structure implied by the CAR and SAR models
journal, April 2004


Space-time modelling of extreme events
journal, September 2013

  • Huser, R.; Davison, A. C.
  • Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 76, Issue 2
  • DOI: 10.1111/rssb.12035

Hierarchical modeling for extreme values observed over space and time
journal, January 2008


Modeling temporal gradients in regionally aggregated California asthma hospitalization data
journal, March 2013

  • Quick, Harrison; Banerjee, Sudipto; Carlin, Bradley P.
  • The Annals of Applied Statistics, Vol. 7, Issue 1
  • DOI: 10.1214/12-AOAS600

Fast sampling of Gaussian Markov random fields
journal, May 2001

  • Rue, Havard
  • Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 63, Issue 2
  • DOI: 10.1111/1467-9868.00288

Stationary max-stable fields associated to negative definite functions
journal, September 2009

  • Kabluchko, Zakhar; Schlather, Martin; de Haan, Laurens
  • The Annals of Probability, Vol. 37, Issue 5
  • DOI: 10.1214/09-AOP455

Spatial process modelling for univariate and multivariate dynamic spatial data
journal, January 2005

  • Gelfand, Alan E.; Banerjee, Sudipto; Gamerman, Dani
  • Environmetrics, Vol. 16, Issue 5
  • DOI: 10.1002/env.715

A Hybrid Pareto Mixture for Conditional Asymmetric Fat-Tailed Distributions
journal, July 2009


Spatiotemporal quantile regression for detecting distributional changes in environmental processes
journal, January 2012


Regression B-spline smoothing in Bayesian disease mapping: with an application to patient safety surveillance
journal, January 2007

  • MacNab, Ying C.; Gustafson, Paul
  • Statistics in Medicine, Vol. 26, Issue 24
  • DOI: 10.1002/sim.2868

Spatial Analyses of Periodontal Data Using Conditionally Autoregressive Priors Having Two Classes of Neighbor Relations
journal, March 2007

  • Reich, Brian J.; Hodges, James S.; Carlin, Bradley P.
  • Journal of the American Statistical Association, Vol. 102, Issue 477
  • DOI: 10.1198/016214506000000753

Assessment of the Impact of Cosmic-Ray-Induced Neutrons on Hardware in the Roadrunner Supercomputer
journal, June 2012

  • Michalak, Sarah E.; DuBois, Andrew J.; Storlie, Curtis B.
  • IEEE Transactions on Device and Materials Reliability, Vol. 12, Issue 2
  • DOI: 10.1109/TDMR.2012.2192736

Classes of Nonseparable, Spatio-Temporal Stationary Covariance Functions
journal, December 1999


Estimating the tail-dependence coefficient: Properties and pitfalls
journal, August 2005


Nonseparable, Stationary Covariance Functions for Space–Time Data
journal, June 2002


A hierarchical max-stable spatial model for extreme precipitation
journal, December 2012

  • Reich, Brian J.; Shaby, Benjamin A.
  • The Annals of Applied Statistics, Vol. 6, Issue 4
  • DOI: 10.1214/12-AOAS591

The t Copula and Related Copulas
journal, April 2005


A hybrid Pareto model for asymmetric fat-tailed data: the univariate case
journal, August 2008


A study of logspline density estimation
journal, November 1991


Efficient inference for spatial extreme value processes associated to log-Gaussian random functions
journal, November 2013


Likelihood-Based Inference for Max-Stable Processes
journal, March 2010

  • Padoan, S. A.; Ribatet, M.; Sisson, S. A.
  • Journal of the American Statistical Association, Vol. 105, Issue 489
  • DOI: 10.1198/jasa.2009.tm08577

Space–Time Covariance Functions
journal, March 2005


A Bayesian Reliability Analysis of Neutron-Induced Errors in High Performance Computing Hardware
journal, June 2013

  • Storlie, Curtis B.; Michalak, Sarah E.; Quinn, Heather M.
  • Journal of the American Statistical Association, Vol. 108, Issue 502
  • DOI: 10.1080/01621459.2013.770694

Hierarchical Spatio-Temporal Mapping of Disease Rates
journal, June 1997

  • Waller, Lance A.; Carlin, Bradley P.; Xia, Hong
  • Journal of the American Statistical Association, Vol. 92, Issue 438
  • DOI: 10.1080/01621459.1997.10474012

An Approach to Statistical Spatial-Temporal Modeling of Meteorological Fields
journal, June 1994


Statistical Modeling of Spatial Extremes
journal, May 2012

  • Davison, A. C.; Padoan, S. A.; Ribatet, M.
  • Statistical Science, Vol. 27, Issue 2
  • DOI: 10.1214/11-STS376

Extreme value analysis for evaluating ozone control strategies
journal, June 2013

  • Reich, Brian; Cooley, Daniel; Foley, Kristen
  • The Annals of Applied Statistics, Vol. 7, Issue 2
  • DOI: 10.1214/13-AOAS628

Dependence modelling for spatial extremes
journal, March 2012


    Works referencing / citing this record:

    Spatiotemporal Modeling of Node Temperatures in Supercomputers [Supplemental Data]
    fileset, June 2016

    • Storlie, Curtis; Reich, Brian; Rust, William
    • figshare-Supplementary information for journal article at DOI: 10.1080/01621459.2016.1195271, 1 MPG file (4.19 MB); 1 RData file (2.42 MB); 1 PDF file (300.99 kB)
    • DOI: 10.6084/m9.figshare.3427406