Measuring and analyzing energy consumption of the data center

Thumbnail Image

Files

master_Makris_Theodosios_2017.pdf (1.82 MB)

URL

Journal Title

Journal ISSN

Volume Title

Sähkötekniikan korkeakoulu | Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.

Authors

Date

2017-08-28

Department

Major/Subject

Radio Communications

Mcode

S3019

Degree programme

CCIS - Master’s Programme in Computer, Communication and Information Sciences (TS2013)

Language

en

Pages

(8) + 39

Series

Abstract

Data centers are continuously expanding, so does the energy consumed to power their infrastructure. Server is the major component of data center’s computer rooms, which runs the most intensive computational workloads and stores the data. Server is responsible for more than a quarter of the total energy consumption of data center. This thesis is focused on analyzing and predicting the energy consumption of the server. Three major components are considered in our study; the processor, the access memory and the network interface controller. We collect data from these components and analyze them using linear regression Lasso model with non-negative coefficients. A power model is proposed for predicting energy consumption at the system-level. The model takes as input CPU cycles and data Translation Lookaside Buffer loads, and predicts the energy consumption of the server with 5.33% median error regardless of its workload.

Description

Supervisor

Xiao, Yu

Thesis advisor

Helske, Juhani

Keywords

data center, energy consumption, system-level power modeling, linear regression

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-201709046832

Collections

[dipl] Sähkötekniikan korkeakoulu / ELEC