A Technique to Minimize the Power Consumption of Computer Fleets

Energy consumption of current computer fleets of most organizations can be substantially reduced by using power management systems. Energy saving is an important matter as, with the sustained increment of the cost of electricity, the monetary savings can be significant. This paper presents a technique to minimize the power consumption of the computer fleet or an organization. The technique is based on a distributed system where a central manager coordinates the local power manager of each computer. A method to determine the profitability of this type of management system is also proposed. The key element to save energy is the idle timer used by local power controllers. When the controller detects that the computer is inactive, it starts the timer, and when the timer expires, the controller puts the computer in suspension or hibernation, or turns off the computer. This paper proposes a systematic technique to estimate the appropriate values of the idle timers for the computers of the fleet. Any administrator of the fleet can take advantage of the technique proposed in this paper to achieve an appropriate trade-off between energy saving and user productivity. The paper also shows the application of the technique to the computer fleet of a company.

[1]  San Murugesan,et al.  Harnessing Green IT: Principles and Practices , 2008, IT Professional.

[2]  Luca Ardito,et al.  Energy Efficiency in the ICT - Profiling Power Consumption in Desktop Computer Systems , 2012 .

[3]  Predrag Buncic,et al.  Software installation and condition data distribution via CernVM File System in ATLAS , 2012 .

[4]  Hongsheng Xi,et al.  Adaptive optimisation of timeout policy for dynamic power management based on semi-Markov control processes , 2010 .

[5]  Luca Benini,et al.  Operating-system directed power reduction , 2000, ISLPED '00.

[6]  Paul Johns,et al.  It's not easy being green: understanding home computer power management , 2009, CHI.

[7]  Marco Mellia,et al.  PoliSave: Efficient power management of Campus PCs , 2010, SoftCOM 2010, 18th International Conference on Software, Telecommunications and Computer Networks.

[8]  Luca Benini,et al.  A survey of design techniques for system-level dynamic power management , 2000, IEEE Trans. Very Large Scale Integr. Syst..

[9]  Richard E. Brown,et al.  After-hours power status of office equipment in the USA , 2005 .

[10]  A. Oxley,et al.  Dynamic PC power state transitioning based on a monitoring service , 2012, 2012 International Conference on Computer & Information Science (ICCIS).

[11]  Wei Liu,et al.  Adaptive power management using reinforcement learning , 2009, 2009 IEEE/ACM International Conference on Computer-Aided Design - Digest of Technical Papers.