Cisco Cisco Energy Management for Distributed Offices White Paper

 

 

© 2014 Cisco and/or its affiliates. All rights reserved. This document is Cisco Public Information. 

Page 1 of 6 

White Paper 

Gaining Energy Transparency and Efficiency in the 
Data Center 

Introduction 

Because of the proliferation of the Internet and information age in the mid 1990s, there has been a rapid expansion 

of data centers. To address the extremely rapid growth of applications and IT services, more and more data 

centers have been built, expanded, and utilized over the last decade and a half. With this growth in data centers, 

energy consumption has also soared to unheard-of levels and shows no signs of slowing. 

According to 

: “The need for backup power 

supplies, environmental controls (air conditioning, fire suppression, etc.) and special security devices mean that 
data centers can consume 200 times as much electricity as standard office spaces.” 

shared that “More than a third (36%) of data center facilities will run out of space, 

power, or cooling, or all the above in 2011 or 2012, according to the Uptime Institute, which recently surveyed 525 

data center operators and owners, with 71% situated in North America. Of those that will run out of one or more 

data center lifelines, 40% plan to build a new data center, 62% plan to consolidate servers, and 29% plan to lease 
collocation space.” 

According to the same Information Week article, “IT is looking to meet growing demands for more computing power 
through consolidation, virtualization, and hardware upgrades.”This evolution in the industry has brought the need to 

be able to measure energy use and effectiveness to the forefront. 

Energy Measurement Challenges 

While the need for accurate energy measurement has long been acknowledged by industry leaders, the tools 

available have been both costly and inadequate. Energy assessments and measurement of electrical loads have 

typically been measured through traditional power meters, measuring actual plug loads. The process of deploying 

meters and measuring individual devices is a very time-consuming and expensive endeavor. According to many 

data center operators, this can even be an extremely risky endeavor since many meters have to be in line to the 

devices they are monitoring and require infrastructure to be taken offline for installation. 

Because of the inability to realistically measure energy on individual data center assets in an efficient way, data 

centers have been unable to measure energy reduction and carbon emissions reductions related to data center 

consolidation, virtualization, and capital equipment upgrade projects. With no way for data centers to measure and 

report the effectiveness of these projects, their success can only be approximated. 

Until now, to read energy on each IT and facilities asset has been a challenge. Only if the asset was instrumented 

for energy or connected to a device that was instrumented for energy could that data be retrieved. Table 1 provides 

some examples of ways to actively measure energy on IT assets and the challenges associated with them.