BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20111115T223000Z
DTEND:20111115T230000Z
LOCATION:TCC 304
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: In this paper, we first study the impact of load placement policies on=0Acooling and maximum data center temperatures in cloud service=0Aproviders that operate multiple geographically distributed data=0Acenters.  Based on this study, we then propose dynamic load=0Adistribution policies that consider all electricity-related costs as=0Awell as transient cooling effects.  Our evaluation studies the ability=0Aof different cooling strategies to handle load spikes, compares the=0Abehaviors of our dynamic cost-aware policies to cost-unaware and=0Astatic policies, and explores the effects of many parameter settings.=0AAmong other interesting results, we demonstrate that (1) our policies=0Acan provide large cost savings, (2) load migration enables savings in=0Amany scenarios, and (3) all electricity-related costs must be=0Aconsidered at the same time for higher and consistent cost savings.
SUMMARY:Reducing Electricity Cost Through Virtual Machine  Placement in High Performance Computing Clouds
PRIORITY:3
END:VEVENT
END:VCALENDAR