Temperature management in data centers

Why some (might) like it hot

Nosayba El-Sayed, Ioan A. Stefanovici, George Amvrosiadis, Andy A. Hwang, Bianca Schroeder

Research output: Chapter in Book/Report/Conference proceedingConference contribution

76 Citations (Scopus)

Abstract

The energy consumed by data centers is starting to make up a significant fraction of the world's energy consumption and carbon emissions. A large fraction of the consumed energy is spent on data center cooling, which has motivated a large body of work on temperature management in data centers. Interestingly, a key aspect of temperature management has not been well understood: controlling the setpoint temperature at which to run a data center's cooling system. Most data centers set their thermostat based on (conservative) suggestions by manufacturers, as there is limited understanding of how higher temperatures will affect the system. At the same time, studies suggest that increasing the temperature setpoint by just one degree could save 2-5% of the energy consumption. This paper provides a multi-faceted study of temperature management in data centers. We use a large collection of field data from different production environments to study the impact of temperature on hardware reliability, including the reliability of the storage subsystem, the memory subsystem and server reliability as a whole. We also use an experimental testbed based on a thermal chamber and a large array of benchmarks to study two other potential issues with higher data center temperatures: the effect on server performance and power. Based on our findings, we make recommendations for temperature management in data centers, that create the potential for saving energy, while limiting negative effects on system reliability and performance.

Original languageEnglish
Title of host publicationSIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems
Pages163-174
Number of pages12
Volume40
Edition1 SPEC. ISS.
DOIs
Publication statusPublished - 13 Aug 2012
Externally publishedYes
Event12th Joint International Conference on Measurement and Modeling of Computer Systems, ACM SIGMETRICS/Performance 2012 - London, United Kingdom
Duration: 11 Jun 201215 Jun 2012

Other

Other12th Joint International Conference on Measurement and Modeling of Computer Systems, ACM SIGMETRICS/Performance 2012
CountryUnited Kingdom
CityLondon
Period11/6/1215/6/12

Fingerprint

Temperature
Servers
Energy utilization
Thermostats
Time and motion study
Testbeds
Cooling systems
Energy conservation
Cooling
Hardware
Data storage equipment
Carbon
Hot Temperature

Keywords

  • CPU
  • data center
  • DRAM
  • energy
  • fans
  • hard drive
  • LSE
  • memory
  • performance
  • reliability
  • temperature

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Networks and Communications

Cite this

El-Sayed, N., Stefanovici, I. A., Amvrosiadis, G., Hwang, A. A., & Schroeder, B. (2012). Temperature management in data centers: Why some (might) like it hot. In SIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems (1 SPEC. ISS. ed., Vol. 40, pp. 163-174) https://doi.org/10.1145/2254756.2254778

Temperature management in data centers : Why some (might) like it hot. / El-Sayed, Nosayba; Stefanovici, Ioan A.; Amvrosiadis, George; Hwang, Andy A.; Schroeder, Bianca.

SIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems. Vol. 40 1 SPEC. ISS. ed. 2012. p. 163-174.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

El-Sayed, N, Stefanovici, IA, Amvrosiadis, G, Hwang, AA & Schroeder, B 2012, Temperature management in data centers: Why some (might) like it hot. in SIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems. 1 SPEC. ISS. edn, vol. 40, pp. 163-174, 12th Joint International Conference on Measurement and Modeling of Computer Systems, ACM SIGMETRICS/Performance 2012, London, United Kingdom, 11/6/12. https://doi.org/10.1145/2254756.2254778
El-Sayed N, Stefanovici IA, Amvrosiadis G, Hwang AA, Schroeder B. Temperature management in data centers: Why some (might) like it hot. In SIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems. 1 SPEC. ISS. ed. Vol. 40. 2012. p. 163-174 https://doi.org/10.1145/2254756.2254778
El-Sayed, Nosayba ; Stefanovici, Ioan A. ; Amvrosiadis, George ; Hwang, Andy A. ; Schroeder, Bianca. / Temperature management in data centers : Why some (might) like it hot. SIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems. Vol. 40 1 SPEC. ISS. ed. 2012. pp. 163-174
@inproceedings{83e76987b7d34d1e9559e75e9ce42563,
title = "Temperature management in data centers: Why some (might) like it hot",
abstract = "The energy consumed by data centers is starting to make up a significant fraction of the world's energy consumption and carbon emissions. A large fraction of the consumed energy is spent on data center cooling, which has motivated a large body of work on temperature management in data centers. Interestingly, a key aspect of temperature management has not been well understood: controlling the setpoint temperature at which to run a data center's cooling system. Most data centers set their thermostat based on (conservative) suggestions by manufacturers, as there is limited understanding of how higher temperatures will affect the system. At the same time, studies suggest that increasing the temperature setpoint by just one degree could save 2-5{\%} of the energy consumption. This paper provides a multi-faceted study of temperature management in data centers. We use a large collection of field data from different production environments to study the impact of temperature on hardware reliability, including the reliability of the storage subsystem, the memory subsystem and server reliability as a whole. We also use an experimental testbed based on a thermal chamber and a large array of benchmarks to study two other potential issues with higher data center temperatures: the effect on server performance and power. Based on our findings, we make recommendations for temperature management in data centers, that create the potential for saving energy, while limiting negative effects on system reliability and performance.",
keywords = "CPU, data center, DRAM, energy, fans, hard drive, LSE, memory, performance, reliability, temperature",
author = "Nosayba El-Sayed and Stefanovici, {Ioan A.} and George Amvrosiadis and Hwang, {Andy A.} and Bianca Schroeder",
year = "2012",
month = "8",
day = "13",
doi = "10.1145/2254756.2254778",
language = "English",
isbn = "9781450310970",
volume = "40",
pages = "163--174",
booktitle = "SIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems",
edition = "1 SPEC. ISS.",

}

TY - GEN

T1 - Temperature management in data centers

T2 - Why some (might) like it hot

AU - El-Sayed, Nosayba

AU - Stefanovici, Ioan A.

AU - Amvrosiadis, George

AU - Hwang, Andy A.

AU - Schroeder, Bianca

PY - 2012/8/13

Y1 - 2012/8/13

N2 - The energy consumed by data centers is starting to make up a significant fraction of the world's energy consumption and carbon emissions. A large fraction of the consumed energy is spent on data center cooling, which has motivated a large body of work on temperature management in data centers. Interestingly, a key aspect of temperature management has not been well understood: controlling the setpoint temperature at which to run a data center's cooling system. Most data centers set their thermostat based on (conservative) suggestions by manufacturers, as there is limited understanding of how higher temperatures will affect the system. At the same time, studies suggest that increasing the temperature setpoint by just one degree could save 2-5% of the energy consumption. This paper provides a multi-faceted study of temperature management in data centers. We use a large collection of field data from different production environments to study the impact of temperature on hardware reliability, including the reliability of the storage subsystem, the memory subsystem and server reliability as a whole. We also use an experimental testbed based on a thermal chamber and a large array of benchmarks to study two other potential issues with higher data center temperatures: the effect on server performance and power. Based on our findings, we make recommendations for temperature management in data centers, that create the potential for saving energy, while limiting negative effects on system reliability and performance.

AB - The energy consumed by data centers is starting to make up a significant fraction of the world's energy consumption and carbon emissions. A large fraction of the consumed energy is spent on data center cooling, which has motivated a large body of work on temperature management in data centers. Interestingly, a key aspect of temperature management has not been well understood: controlling the setpoint temperature at which to run a data center's cooling system. Most data centers set their thermostat based on (conservative) suggestions by manufacturers, as there is limited understanding of how higher temperatures will affect the system. At the same time, studies suggest that increasing the temperature setpoint by just one degree could save 2-5% of the energy consumption. This paper provides a multi-faceted study of temperature management in data centers. We use a large collection of field data from different production environments to study the impact of temperature on hardware reliability, including the reliability of the storage subsystem, the memory subsystem and server reliability as a whole. We also use an experimental testbed based on a thermal chamber and a large array of benchmarks to study two other potential issues with higher data center temperatures: the effect on server performance and power. Based on our findings, we make recommendations for temperature management in data centers, that create the potential for saving energy, while limiting negative effects on system reliability and performance.

KW - CPU

KW - data center

KW - DRAM

KW - energy

KW - fans

KW - hard drive

KW - LSE

KW - memory

KW - performance

KW - reliability

KW - temperature

UR - http://www.scopus.com/inward/record.url?scp=84864718143&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84864718143&partnerID=8YFLogxK

U2 - 10.1145/2254756.2254778

DO - 10.1145/2254756.2254778

M3 - Conference contribution

SN - 9781450310970

VL - 40

SP - 163

EP - 174

BT - SIGMETRICS/Performance 2012 - Proceedings of the 2012 ACM SIGMETRICS/Performance, Joint International Conference on Measurement and Modeling of Computer Systems

ER -