This ap proach is based on the assumption that the mechan ics of a drive dominate the electronics in terms of power consumed Our 64257negrained measurements in dicate a more complex situation We present results which show the importance of a detaile ID: 25430
Download Pdf The PPT/PDF document "Hard Drive Power Consumption Uncovered C..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
HardDrivePowerConsumptionUncovered ComputerLaboratoryDigitalTechnologyGroup AnthonyHylick,AndrewRice,BrianJones,RipdumanSohanMotivation Attemptstoreducepowerconsumptionhavemainlyfocusedonmaximizingstandbyperiods.Thisap-proachisbasedontheassumptionthatthemechan-icsofadrivedominatetheelectronicsintermsofpowerconsumed.Our ne-grainedmeasurementsin-dicateamorecomplexsituation.Wepresentresultswhichshowtheimportanceofadetailedunderstand-ingofpowerconsumptionandidentifytheneedforamoreexpressiveAPIbetweentheOSandhardwaredevicestomaximizepowereciency. Table1.Drivedetailsandenergyconsumption ExperimentalSetup&Results Wehavedevelopedameasurementplatform(Figure1)toprovidedirect,onlinemeasurementsofindivid-ualPCcomponents.Theplatformprovidestwenty-four,12-bitinputs,eachsampledat98Hz.Wehavebegunourinvestigationwithharddrivesandexam-inedthreedrivestates|read,write,andstandby.Closeattentionwasalsogiventothesigni canttran-sitionsbetweenthem(Figure2).ResultsfortesteddrivesaresummarizedinTable1,andFigure3isanannotated,graphicalrepresentationofthepowerconsumptionofarepresentativedrive(Drive6)dur-ingarunofourtestscript. Figure1.Measurementenvironment Figure2.TestSequenceforFigure3RegionAandFDevicenegotiationquietlyconsumesasigni cantamountofpowereventhoughnoworkonthediskisbeingdone. Figure3.PowerConsumedbyDrive6RegionITransitioningtostandbysigni cantlyreducesthepowerconsumption,however,theelectronicscontinuetoconsume1.5W.RegionCandGThepowerconsumptionofreadsandwritesarenoticeablyasymmetric.For1GBofdata,readscostatleasttwiceasmuchaswritesin75%ofourtesteddrives.RegionJTheamountofenergyrequiredtospin-upadriveis xed,regardlessoftheamountoftimespentinstandby. R/WAsymmetryNewervs.OlderDrivesMoreExpressiveAPI-readscostmorethanwrites-newermoreenergyecient-powermodelismorecomplex-newpower-savingpossibilities-capacityandRPMpropelindustry-alldrivesaredi erent(whataboutpowereciency?)-moreexpressiveAPIisneededhttp://www.cl.cam.ac.uk/Research/DTG