Department - Author 1

Computer Science Department

Degree Name - Author 1

BS in Computer Science

Date

6-2010

Primary Advisor

Phillip Nico

Abstract/Summary

The purpose of this project is to demonstrate the idea that clustering together less powerful but energy efficient machines that are commercially available can provide more efficient computing. With a third of energy consumption in the United States attributed to data centers, power efficiency has become an increasingly important topic in computing. Companies like Google have exceeded their grid’s capacities and built datacenters next to hydroelectric dams to reduce power loss from transmission and to increase their data center’s capacity. These experiments aspire to demonstrate a model for achieving a higher performance-per-watt in a larger scale using commercially available low-power hardware.

Share

COinS