Software or Hardware: The Future of Green Enterprise Computing

Over the last few years, interest in “green computing” has motivated research into energy-saving techniques for enterprise systems, from network proxies and virtual machine migration to the return of thin clients. This paper tries to answer a possibly contentious question: would we be better served by the embarrassingly simple approach of replacing every desktop with a laptop? To answer this question, we use power and utilization data collected from more than 100 devices over durations up to 15 months. We find that choosing the right computing systems – laptops – would save more energy than stateof-the-art power management software or thin clients. Furthermore, the marginal savings of applying software techniques on top of laptops is so small that it is probably not worth the trouble. When selecting computers, there are many other considerations than just energy, such as computational resources, and price. We find that these factors generally do not reduce the attractiveness of a laptop-based enterprise. We discuss current trends in enterprises today, and how our conclusions might affect their directions, sketching a future of how a cost-efficient enterprise might become a hybrid system entwining laptops and tablets with a computing cloud.

[1]  Philip Levis,et al.  Identifying Energy Waste through Dense Power Sensing and Utilization Monitoring , 2010 .

[2]  Renata Teixeira,et al.  Explaining packet delays under virtualization , 2011, CCRV.

[3]  Jason Nieh,et al.  THINC: a virtual display architecture for thin-client computing , 2005, SOSP '05.

[4]  Janardhan R. Iyengar,et al.  Low Extra Delay Background Transport (LEDBAT) , 2012, RFC.

[5]  Adrian Schüpbach,et al.  The multikernel: a new OS architecture for scalable multicore systems , 2009, SOSP '09.

[6]  Kang G. Shin,et al.  LiteGreen: Saving Energy in Networked Desktops Using Virtualization , 2010, USENIX Annual Technical Conference.

[7]  Junda Liu,et al.  Skilled in the Art of Being Idle: Reducing Energy Waste in Networked Systems , 2009, NSDI.

[8]  G. T. Gardner,et al.  The Short List: The Most Effective Actions U.S. Households Can Take to Curb Climate Change , 2008 .

[9]  Guido Appenzeller,et al.  Sizing router buffers , 2004, SIGCOMM '04.

[10]  Aman Kansal,et al.  Sleepless in Seattle No Longer , 2010, USENIX Annual Technical Conference.

[11]  Rajesh Gupta,et al.  SleepServer: A Software-Only Approach for Reducing the Energy Consumption of PCs within Enterprise Environments , 2010, USENIX Annual Technical Conference.

[12]  John F. Busch,et al.  After-hours Power Status of Office Equipment and Inventory of Miscellaneous Plug-load Equipment - eScholarship , 2004 .

[13]  Jonathan W. Hui,et al.  T 2 : A Second Generation OS For Embedded Sensor Networks , 2005 .

[14]  David E. Culler,et al.  Design and implementation of a high-fidelity AC metering network , 2009, 2009 International Conference on Information Processing in Sensor Networks.

[15]  David E. Culler,et al.  The dynamic behavior of a data dissemination protocol for network programming at scale , 2004, SenSys '04.

[16]  Eyal de Lara,et al.  The Case for Energy-Oriented Partial Desktop Migration , 2010, HotCloud.

[17]  Paramvir Bahl,et al.  Somniloquy: Augmenting Network Interfaces to Reduce PC Energy Usage , 2009, NSDI.

[18]  Philip Levis,et al.  Collection tree protocol , 2009, SenSys '09.