You are on page 1of 4

2011 IEEE/ACM International Conference on Green Computing and Communications

The Survey and Future Evolution of Green Computing

Qilin Li

Mingtian Zhou

Production and Technology Department


Sichuan Electric Power Science and Research Institute
Chengdu, Sichuan, P.R.China
li_qi_lin@163.com

School of Computer Science and Technologies


University of Electronic Science and Technology
Chengdu, Sichuan, P.R.China
mtzhou@uestc.edu.cn
Although green computing is becoming increasingly
important in IT systems, it presents challenging problems to
system designers. Designers need to take into account energy
consumption during the phase of system design and to find
solutions to reduce it. Green computing involves all aspects
of IT systems, such as chips, system architectures, compilers,
operating systems, communication networks and application
services, and so on[1]. Further, these aspects are
interdependent and complicated. As a result, building lowcost and low-power-consumption systems is a challenging
and important activity. Such a new computing paradigm
introduces new technical challenges to system designers.
In light of todays requirements for green computing, we
present latest research efforts that attempt to deal with them
and indicate still open issues. We thus discuss the
connotation of green computing and sketch our view on the
next generation of IT systems for green computing. We
further identify key issues relevant to green computing and
evaluate different approaches to these problems. The scope
of this survey covers basic concept as well as techniques and
methods in support of green computing.
The remainder of this paper is organized as follows. In
section 2, we introduce the connotation of green computing.
In section 3, we look deeper into some key issues relevant to
green computing and some real solutions are shown as well.
Finally, section 4 points out future directions of research in
the area of green computing and concludes with the paper.

AbstractThe concept of green computing has begun to spread


in the past few years, gaining increasing popularity. Besides
the widespread sensitivity to ecological issues, such interest
also stems from economic needs, since both energy costs and
electrical requirements of IT industry around the world show a
continuously growing trend. In this paper, we first discuss the
connotation of green computing and sketch our view on the
next generation of IT systems for green computing.
Subsequently, we identify key issues relevant to green
computing and evaluate different approaches to these
problems. Finally, we point out future directions of research
and conclude the paper.
Keywords- Green Computing; Low Power;
Conusmption; Energy-Aware; Green Networking

I.

Energy

INTRODUCTION

Over the last decade, IT Technologies are becoming


increasingly prevalent due to the explosive growth of the
Internet and decreasing cost of computer hardware. In
consequence, a variety of applications from different
domains such as those in the telecommunications industry,
the medical field and the financial sector are relying much
more heavily on services provided by IT technologies. IT
Technologies have been the foundation for the rapid
development of society and economy. As human beings
increasingly depends on the IT systems, the problem of
energy consumption in IT industry becomes more acute. The
fact has made it necessary to investigate a new paradigm of
computing, called green computing.
Green computing refers to environmentally sustainable
computing. Green computing attempts to reduce the cost and
power consumption of IT system and maximize energy
efficiency during the systems lifetime. The main purpose of
green computing is to investigate new computer systems,
computing model and applications with the low-cost and
low-power-consumption and promote the sustainable
development of economy and society[1][2]. Specifically, it
entails system architecturesystem softwareparallel and
distributed computing and computer network[1]Nowadays
green computing has been one of major trends in the IT
industry and is drawing much attention from the government
and academic research and engineering practice.

978-0-7695-4466-3/11 $26.00 2011 IEEE


DOI 10.1109/GreenCom.2011.47

II.

THE CONNOTATION OF GREEN COMPUTING

Although green computing is considered one of the most


promising technology to date in IT industry and is gaining
wide-spread popularity in IT ocmpanies, research institutes
and the government. However, the concept of green
computing in the field lacks universally accepted definitions.
In general terms, green computing is defined as the study
and practice of designing, manufacturing, using, and
disposing of computers, servers, and associated
subsystemssuch as monitors, printers, storage devices, and
networking and communications systems efficiently and
effectively with minimal or no impact on the
environment[1][2]. Green computing is the environmentally
responsible use of computers and related resources. Such
practices include the implementation of energy-efficient
central processing units (CPUs), servers and peripherals as
224
231
230

well as reduced resource consumption and proper disposal of


electronic waste (e-waste) [1][2].
As a computing field, green computing is a global
concept that entails system architecture, system software,
parallel and distributed computing and computer network. It
aims to reduce the power consumption of computer systems,
provide high-efficient, dependable and pervasive services,
and achieve the objective of low power of IT systems[1][2].
The energy comsumption of computer system is
dertermined by the power consumption of hardware, runtime efficiency of tasks and configuration policy of
resources[1]. Accordingly, green computing attempts to
construct a new energy-awareness computing system, internetworking environment and computing service facilities,
supports the new kind of computer architecture and
computing paradigm like cloud computing, and provides a
low-power computing environment for personalizd and
diversified information service[1][2]. With rapid advances in
system performance and reliability, green computing can
reasonably exploit computing resource and meet the
increasingly diversified computing requirements in a lowpower manner[1][2].
III.

an overall power consumption model for computer system


still need further investigation.
In order to measure and evaluation the power
consumption, several ways have been proposed[1]. One way
is hardware measurement that uses instruments to obtain
real-time voltage and current and compute the power and
consumption. Hardware measurement requires the support of
special hardware. Another way is theory computing. Theory
computing needs to exploit working frequency and voltage
to calculate the power consumption of CPU. Besides, it
computes the power consumpiton of peripherals according to
the current state. Theory computing lacks in accuracy. A
third alternative is software simulation. Examples of systems
built around the principle of software simulation include the
Wattch system[6] and Simplepower system[7]. The major
limitations of these software systems just obtain the coarse
power consumption at the level of nodes.
B. Energy-awareness
Energy-awareness allows computer systems to run in a
low power-consumption manner under a broader range of
conditions. By means of energy-awareness, a computer
system can tune the power-consumption states of system
components and reduce the system energy consumption . In
order to support energy-awareness, computer system needs
to collect the run-time information of applications, monitor
the power-consumption states of system components,
compute energy-awareness decisions, and notify applications
about states changes. Energy-awareness strategies involve
techniques in how systems and applications respond to the
state changes and ajdust the behaviors of system
components.
The range for energy-awareness strategies is categorized
into two extremes. At one extreme, energy-awareness is
entirely the responsibility of individual applications. This
approach allows the applications to adjust the voltages
supplied to the hardware, which reduces both the amount of
heat produced and electricity consumed. While this strategy
avoids the need for system support, it makes energy-aware
applications more difficult to write.
At the other extreme, system strategy places the whole
responsibility for energy-awareness on the computer system.
This approach does not require changing existing
applications. It allows computer system to directly control
the power-saving aspects of its underlying hardware. As
such, a system can automatically turn off components such
as monitors and hard drives after set periods of inactivity. In
addition, a system may hibernate, where most components
(including the CPU and the system RAM) are turned off.
However, the intrinsic limitation of system strategy is
that there may be situations in which the energy-awareness
performed by the system is adequate or even
counterproductive. It could sacrifice performance and
functionality of a computer system. As the applications are
shielded from dealing with enery-awareness, it might be very
difficult for the system to make adaptive decisions that meet
the needs of different applications.
Between these two extremes are defined as applicationaware strategy. This approach supports collaborative energy-

KEY ISSUES

A. The Modeling and Evaluation of Energy Efficiency


Energy efficiency is an important aspect of green
computing. In order to achieve green computing, it is
essential for researchers and system practioners to
investigate the relationship between energy efficiency and
other impact factors and construct the theory model and its
evaluation mechanism of energy efficiency. The problem of
energy efficiency entails all aspects of a computer systems,
including chips, system architecture, compilers, operating
systems, communication network, and application and so
on[1].
There are two compelling reasons for in-depth study of
the power usage model of Todays computing system. First
of all, there is need for a scientific approach to quantify the
energy cost of typical computing systems[3][4][5]. Such cost
estimates could be used to accurately estimate future
machine operation costs for common application types.
Second, an overall study may reveal opportunities for power
and energy savings.
Previous studies of the modeling of power consumption
focus on one particular aspect or system[3][4][5]. Such studies
separate individual systems or components and do not take
into account all aspect of computer systems, such as
hardware platform, system software and applications. For
example, T. K. Tan[3] investigates the problem of modeling
of power consumption in embedded systems , and presents a
systematic methodology to perform energy analysis and
macro-modeling of an embedded OS. Other attempts[4][5] to
estimate power consumption for systems such as scientific
computing and parallel computing in distributed systems.
Obviously, These approaches could be completely inaccurate
for present computing systems as power usage increases
exponentially for some components. As a result, constructing

231
232
225

awareness between the applications and the system. In more


detail, the applications can decide how to best adapt to the
dynamically changing system run-time states while the
system monitors the power-consumption states of its
hardware, collects valuable information, notifies application
of relevant changes and enforces low power-consumption
policies. Some representative techniques built around the
principle of energy-awareness include compiler optimization
[8][9]
and dynamic power consumption management of
operating systems[10][11].
Although there have been many research efforts to
reduce the energy usage of specific hardware devices and
contruct an energy-awareness system, few systems consider
energy as a first-class systems resources that cuts across all
existing system resources, such as CPU, disk, memory, and
the network in a unified manner[12]. As a result, many issues
requires further investigations and there are groups of
researchers that are already working to find an unifying
abstraction to integrate energy-awareness into tasks
management of the operating system, hardware abstraction
and resource allocation[12]. This allows the energy-awareness
computer system to reason about the overall energy behavior
of an application on a platform-specific basis and to
potentially extend the useful lifetime of the system for
unmodified applications[12].

network/device portions to low standby modes, and to wake


them up only if necessary.
Conventional fixed networks environment is always
based on the assumptions that the physical location of
services and the connection among services do not change
during the computation. However, in todays networking
environment, these assumptions are rarely valid or
appropriate. Thus, sleeping/standby strategies have to be
explicitly supported with special proxying techniques that
are able to maintain the network presence of sleeping
nodes[13]. More specifically, an enhancement for
sleeping/standby solution is to maintain continuous network
presence which consists of having a network host transfer
network presence to a proxy, namely Network Connectivity
Proxy (NCP), when entering sleep mode[13]. The main
purpose of such proxy is to respond to routine network traffic
as the device sleeps, and to wake the device up only when it
is truly necessary[13].
IV.

CONCLUSION AND FUTURE WORK

The concept of green computing has begun to spread in


the past few years, gaining increasing popularity. Besides the
widespread sensitivity to ecological issues, such interest also
stems from economic needs, since both energy costs and
electrical requirements of IT industry around the world show
a continuously growing trend.
This paper aimed at providing an up-to-date survey on
the current state-of-the-art in green computing. To this aim,
we firstly introduce the connotation of green computing.
Subsequently, we provide a detailed survey on some key
issues of green computing, including the modeling and
evaluation of energy efficiency, energy-awareness and green
networking. In addition, details of some real solutions have
been showed as well.
Although current studies on green computing have
addressing some issues, many issues require further
investigation. A first issue[1] that needs to be addressed is
how to provide the solutions for large-scale computing
system like data centers and super computer centers where
applications may have large energy consumption. Another
issue[1] concerns an overall energy-saving study for computer
systems. The existing approaches focus on one particular
aspect of the computing system and few studies consider
energy consumption as an overall problem to solve in an
integrated manner. Other open issues[13] include green
metrics and standard benchmarking technologies for network
devices, the definition of green data and control plain
abstraction layers; green support and management of
redundant devices; virtualization of networks and
networking equipment.

C. Green Networking
As the new generation networking environment is taking
shape, energy efficiency has pervaded the network
infrastructure as a whole, to such extent as to become part of
the network design criteria, and to carry across multiple
networking domains for the achievement of a general target.
A common opinion among green networking community is
that the sole introduction of low consumption silicon
technologies may not be enough to for drawing ahead current
network equipment towards a greener networking
environment. As a consequence, The problem of energy
efficiency of networking environment is a concern as in
other green computing areas.
From a general point of view, there are three major
approaches to green networking. One strategy for achieving
this is re-engineering. Re-engineering approaches devote to
introducing and designing more energy-efficient components
for network device architectures, at suitably dimensioning
and optimizing internal organization devices, as well as at
reducing their intrinsic complexity levels[13]. Some novel
energies-efficent networking technologies includes new
silicon and memory technologies, and media/interface
technologies for network links[13].As another example,
dynamic adaptation approaches aim at modulating capacities
of network device resources, such as link bandwidths,
computational capacities of packet processing engines, and
so on, according to current traffic loads and service
requirements[13]. This can be performed by using two power
aware capabilities, namely, dynamic voltage scaling and idle
logic, which both allow the dynamic trade-off between
packet service performance and power consumption[13]. The
other strategy is sleeping/standby. Sleeping/standby
approaches are used to smartly and selectively drive unused

ACKNOWLEDGMENT
The authors would like to acknowledge the anonymous
reviewers. This work has been partially funded by National
11-5th
High-Tech
Support
Program
of
China
(2006BAH02A0407).

232
233
226

REFERENCES
[1]
[2]

[3]

[4]

[5]

[6]

[7]

[8]

Guo Min yi, Green Computing Connotation and Tendency, in


Computer Engineering, 36(10):pp.1-7, May 2010
Bin Guo, Yan Shen, Zili Shao, The redefinition and some discussion
of Green Computing, in Journal of Computer, 32(12):pp.1-14,
Dec.2009.
Tan T K Raghunathan A Jim N K, Energy Macro-modeling of
Embedded Operating Systems, ACM Transactions on Embedded
Computing Systems20054(1)231254.
Feng Xizhou, Ge Rong, Cameron K w, Power and Energy Profiling
of Scientific Applications on Distributed Systems, ProcOf IPDPS
05, Denver, ColoradoUSAIEEE Press2005.
Bni VNorris BHuck K et a1, A Component InfrasU'ueture for
Performance and Power Modeling of Parallel Scientific Applications,
Procof CBI-IPC08, KarlsruheGermany2008.
Brooks D Tiwari V Martonosi M Wattch A Framework for
Architectural-level Power Analysis and Optimizations, Proc Of
ISCA'00Vancouver,CanadaACM Press2000.
Ye w, Vijaykrishnan NKandemir Met a1, The Design and Use
of SimplepowerA Cycle-accurate Energy Estimation Tool, Proc of
the 37th Annual Design Automation Conference, New York, USA,
ACM PTess2000.

[9]

[10]

[11]

[12]

[13]

233
234
227

Huang Po-KuanGhiasi S, Efficient and Scalable Compilerdirected


Energy Optimization for Realtime Applications, ACM Transactions
on Design Automation of Electronic Systems, 12(3)pp.1-16,
2007.
Shao zili, High Performance Low Power and Secure Embedded
Systems[D], Texas Dallas USA School of Engineering and
Computer ScienceUniversity ofTexas at Dallas2005.
Burd T D Brodersen R w, Design Issues for Dynamic Voltage
Scaling, Proc of the International Symposium on Low Power
Electronics and Design, RapalloItalyACM Press2000.
Benini L Micheli G Dynamic Power Management Design
Techniques and CAD Tools[M], Norwell MA USA Kluwcr
Academic Publishers1998.
Zeng HengEllis C SLebeck AIt,et a1, ECOSystemManaging
Energy As a First Class Operating System Resource, ACM SIGPLAN
Notices, 200237(10)pp.123-132.
Bolla, R., Bruschi, R., Davoli, F., Cucchietti, F., Energy Efficiency in
the Future Internet: A Survey of Existing Approaches and Trends in
Energy-Aware Fixed Network Infrastructures, Communications
Surveys & Tutorials, IEEE Issues, pp(99): 1-22.

You might also like