DocumentCode
114119
Title
System architecture support for green enterprise computing
Author
Kazandjieva, Maria ; Shah, Chinmayee ; Cheslack-Postava, Ewen ; Mistree, Behram ; Levis, Philip
fYear
2014
fDate
3-5 Nov. 2014
Firstpage
1
Lastpage
10
Abstract
This paper proposes a novel system for enterprise computing that reduces energy consumption without sacrificing performance or putting devices to sleep. It uses a hybrid architecture composed of multiple device classes and it runs an application in the most energy-efficient location. Our prototype, called Any-ware, provides desktop-class performance while reducing energy consumption by 75% through a combination of lightweight clients and a small number of servers. Providing such an elastic system invisibly to end users requires solving many challenges, including deciding where to run applications while simultaneously making it appear they all run locally. Anyware´s hybrid design suggests a new way to think about using the modern spectrum of personal computing devices.
Keywords
business data processing; green computing; power aware computing; Any-ware; desktop-class performance; energy consumption reduction; energy-efficient location; green enterprise computing; hybrid architecture; personal computing devices; system architecture support; Computer architecture; Energy consumption; Portable computers; Predictive models; Program processors; Servers; energy efficiency; hybrid computing; system architecture;
fLanguage
English
Publisher
ieee
Conference_Titel
Green Computing Conference (IGCC), 2014 International
Conference_Location
Dallas, TX
Type
conf
DOI
10.1109/IGCC.2014.7039143
Filename
7039143
Link To Document