‧Givesstudentsperspectivesonhowsharedmemoryparallelprogramsarewritten
‧Providesanunderstandingofwhatsoftwareprimitivesandhardwaresupportareneededinsharedmemorymulticoreandmultiprocessors
‧Discussesmemoryhierarchy,basicproblemswhendesigningsharedmemoryparallelmultiprocessors,cachecoherence,memoryconsistency,synchronization,andinterconnectionnetwork,showingstudentshowtheconceptsinteractandfittogether
‧Exploresthesingle-instructionmultiple-thread(SIMT)programmingmodel,whichisoftenusedinthegraphicsprocessingunit(GPU)system
‧Includesend-of-chapterexercises,withsolutionsandothermaterialonthebook’sCRCPresswebpage
Althoughmulticoreisnowamainstreamarchitecture,therearefewtextbooksthatcoverparallelmulticorearchitectures.Fillingthisgap,FundamentalsofParallelMulticoreArchitectureprovidesallthematerialforagraduateorseniorundergraduatecoursethatfocusesonthearchitectureofmulticoreprocessors.Thebookisalsousefulasareferenceforprofessionalswhodealwithprogrammingonmulticoreordesigningmulticorechips.
Thetext’scoverageoffundamentaltopicspreparesstudentstostudyresearchpapersinthemulticorearchitecturearea.Thetextoffersmanypedagogicalfeatures,including:
‧Sufficientlyshortchaptersthatcanbecomfortablyreadoveraweekend
‧Introducingeachconceptbyfirstdescribingtheproblemandbuildingintuitionthatleadstotheneedfortheconcept
‧““Didyouknow?““boxesthatpresentminicasestudies,alternativepointsofview,examples,andotherinterestingfactsordiscussionitems
‧Thought-provokinginterviewswithexpertswhosharetheirperspectivesonmulticorearchitecturesinthepast,present,andfuture
‧Onlineprogrammingassignmentsandsolutionsthatenhancestudents’understanding
Thefirstseveralchaptersaddressprogrammingissuesinsharedmemorymultiprocessors,suchastheprogrammingmodelandtechniquestoparallelizeregularandirregularapplications.Thecoreofthebookcoversthearchitecturesforsharedmemorymultiprocessors.Thefinalchaptercontainsinterviewswithexpertsinparallelmulticorearchitecture.
GPU PARALLEL ARCHITECTURE SYSTEM
