支持硬件任務(wù)可搶占的CPU-FPGA混合架構(gòu)的軟硬件任務(wù)遷移研究.pdf

1、浙江大學(xué)計(jì)算機(jī)科學(xué)與技術(shù)學(xué)院碩士學(xué)位論文支持硬件任務(wù)可搶占的CPUFPGA混合架構(gòu)的軟硬件任務(wù)遷移研究姓名：馮德貴申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：陳天洲施青松20100101AbstractAstheCPUhasgoodflexibilitybutpoorperformance，thereconfigurabledevicehasbetterperformancebutwithlongconfiguredelaytheC

2、PU／FPGAhybridarchitecnlrebecameoneofthemostimportantformsTheresearchaboutthescheduleabouthybridtasksisahotspotThescheduleapproachesabouthybridtaskscanclassifyintotwoclassificationsOneisaboutnonpreemptivemultitasking，theo

3、therispreemptivemultitaskingForthesecondclassification，asitneedsthearchitecturesupporttheresearchisfewerThispaperpresentsthearchitecturewhichsupportpreemptivemultitaskingForeachreconfigureregion，anintermediatedatabufferi

4、saddedtOstorethestateinformationduringpreemptingAreconfiguremigrationmanagerisaddedfortheentirereconfiguredevicetomanagethereconfigureregionsforpreemptThispaperpresentsanalgorithmtoselecttheappropriatemigrationpointAsthe

5、migrationintroducedthetraditionalschedulealgorithmsarenotsuitable，thispaperpresentanewschedulealgorithmbasedmigrationAsthetaskcanmigratefromoneplacetoanotherthenthetraditionalcommunicationbasedNoCbetweenthemmayerrorTodea

6、lwiththesituation，thispaperalsointroducesandedicatednamingnodefortransformationbetweenthenodeidandnodenameExperimentalresultsshowthatthealgorithmbasedmigrationincreasesthe1294％acceptrateforhardwaretasksSothearchitecturet