SIA：美國半導體研究：通過創新獲得領導力

AmericanSemiconductorResearch:ndglobalderntechnologiespowerincredibleveinternationalcompetitionLearnmoreatwwwsemiconductorsorgblershipfulinnovationstemtofsemiconductorsTheU.S.canemployarangeofthebestpracticestoensurethesuccessofthispublicinvestmentAppendix1arsagobetsisrequiredforinnovationstovolumeproductionR&DisacriticalpartofavirtuouscycleofinnovationthatsupportsU.S.technologyleadership.Innovationsyieldsuperiortechnologiesandproductsthat,whenusedincommercialproduction,providethefundsneededtomakemassiveinvestmentsinfutureR&D.Todeveloptheseinnovations,theU.S.semiconductorindustryinvested$50billioninR&Din2021alone.Withthepassageofthe2022CHIPSandScienceact,thefederalgovernmentispoisedtomakeitssinglelargestinvestmentinsemiconductorR&Dever.AnationalstrategyforsemiconductorR&DshouldtargetthisinvestmenttowardcriticalgapsintheU.S.R&Decosystemtorevitalizetheinnovationpipeline,toalignR&Dwithcommercialpriorities,andtostrengthenU.S.technologicalcompetitiveness.esrntsDincreasedomesticsemiconductorcapacity.keyeasnductershipntheirrespectivemissions2logydevelopmentcapabilitiesnstthebenefitsofitimportantforpilotingandprototypingtoaccelerateationeffortsThewillrcheffortsistheestablishmentoftransitionpathforpromisingtechnologiesthroughscaleupcomplextechnologicalproblemsthatbenefitfromcollaborationacrossthefullcomputingstackgneweizationseersetechnologies.eeconomiccompetitivenessitaceofinnovation1.The2030DecadalPlanforSemiconductorsidentifies8highlevelcomponentsofthecomputestack:applications,software,algorithms,architectures,circuits,devices,structures,andmaterials.2.SeeBCGxSIA’sStrengtheningtheGlobalSemiconductorValueChainforadditionalinformationonthesevendifferentiatedactivitiesofthesemiconductorvaluechain3Semiconductors,orchips,arepervasivein,andincreasinglycriticalto,thefunctioningofthemodernworld.Frompoweringdatacentersanalyzinghistoricallyunprecedentedquantitiesofdata,tocontrollingtheMarsroverPerseveranceinchallengingenvironments,theworlddemandsmorefromitssemiconductorstodaythanitdidjustafewyearsago.Meetingthesedemands–andenablinginnovationatsemiconductor-dependentcompaniesacrosstheU.S.economy-willrequiresustainedinvestmentsinsemiconductorresearchanddevelopment(R&D).rdesignorduringsemiconductorsreportwilldiscusstheroleofcompetitiveness.inimportantarch4retionsratherthanamongthemroductionalinnovationsisrequiredforivencturelimitationsmanypotentialinnovationsoftenciatedwithscalingcanbedaunting5eRDinnovationsthroughoutthisprocess.performancebeyondthelimitsdiontoadvancedsemiconductormaterialsthroughthephasesofmaterialsforuseinarangeoffromelectricvehiclestomobilenetworksdefenseostimportantercializedinthescanwindupacceleratingeffortsofotherversariestainFortunatelycontainopportunitiestostrengthenUltravioletLithographytheU.S.R&Decosystemandextendthevirtuouscycleofinnovation–andU.SUltravioletLithographyAsanexampleofaninnovationthatbeganheUSbutwascommercializedrseasconsiderextremeultravioletEUVlithography.Lithographyisaprocessinsemiconductormanufacturingthatuseslighttoproduceextremelysmallpatternsonmaterialslikesiliconwafers.EUVlithographyisahighlyadvancedversionofthisprocessusedformanyofthemostadvancedchips,likethosefoundinleadingsmartphones.U.S.publicinvestmentplayedanearlyAsanexampleofaninnovationthatbeganheUSbutwascommercializedrseasconsiderextremeultravioletEUVlithography.Lithographyisaprocessinsemiconductormanufacturingthatuseslighttoproduceextremelysmallpatternsonmaterialslikesiliconwafers.EUVlithographyisahighlyadvancedversionofthisprocessusedformanyofthemostadvancedchips,likethosefoundinleadingsmartphones.U.S.publicinvestmentplayedanearlyandsustainedroleinsupportingEUVlithography,whileparallelinvestmentsweremadeinKoreaandJapan.Newapplicationsdrovetheneedforcontinuedtransistorscalinginthelate1980s.WhileEUV’spotentialwasknown,manyintheindustryconsideredEUVinfeasiblegiventechnicalfundedtheAdvancedLithographyProgramwhichconductedearlyresearchintoEUVMATECHanotforprofitconsortiumthatperformedR&Dtoadvancechipmanufacturing,partneredacrossindustryandacademiaforover15yearstoaccessandbuildoutinfrastructureandexpertisethatindustryaloneconsideredtoorisky.Acollectionofcompanies,includingASML,Intel,Samsung,andTaiwanSemiconductorManufacturingMCinvestedbetweenbilliontomatureEUVintoacommerciallyviabletechnology.ASMLtodayistheonlycompanywiththeabilitytoimplementEUVlithographytechnologiesinacommerciallyviableway.AsgandIntelareusingEUVtodevelopsemiconductorprocesstechnology.thestrategiesforimprovingcomputationtechnologiescomingyearsrativedevelopmentacrossthecomputingstack.Semiconductorshavedriventransformativeadvancesineverymoderntechnology.Chipsalsowillunderpinadvancesinthe“must-win”technologiesofthefuture-includingartificialintelligence(AI),quantumcomputing,andadvancedwirelessnetworks–makingcontinuedU.S.leadershipinsemiconductorscriticaltoourfuture.3AnationalstrategyforsemiconductorR&D,focusedonrevitalizingthepipelineofinnovationandaligningitwithcommercialpriorities,willaddresscriticalgapsintheU.S.R&DecosystemandstrengthenU.S.technologicalcompetitivenessandleadershipthroughthefollowingdecades.67ndenrtainortoodifficultforasinglefirmtoturnsforpotentialsemiconductorinnovations.heademiaplaysasignificantstartupsinindustry.rformdesignandeQualcomm,AMD,NVIDIA,andApple)focusonsemiconductordesign,thehighestvalue-addingndpartnerwithothercompaniesforsemiconductorfabrication.9?FoundriesmeetthefabricationneedsofbothIDMsandfablessdesignfirms.FoundriesoftenfocusexclusivelyonfabricatingchipsdesignedbyothersegTSMCSamsungFoundry,IntelFoundryServices,andGlobalFoundries).FoundriesinvestinR&Drelatedtomanufacturingtechnologiesandareresponsibleforalargeshareofadvancedmanufacturing.10conductorequipmentandmaterialscompanies(e.g.,AppliedMaterials,ASML,EMDElectronics,LamResearch,TokyoElectron)developadvancedprocessandinspectionsystemsrequiredtoproducechips.EquipmentandmaterialscompaniesinvestinR&Dfornewprocesstechnologies,materialsandmanufacturingmethodsthatenablethesemiconductortechnologyroadmap,andcontributesignificantamounttoR&DinvestmentinthesemiconductorindustryonparwithseveralU.S.chipmanufacturers.tor(FFRDCs),whichincludethenationallabs,arepublic-privatepartnershipsoperatedbyacademicinstitutionsorprivatecompaniesthatarefundedbyandperformresearchforthefederalgovernment.iessgeofapplicationsfromartificialintelligencetoautomotiveenceandEngineeringIndicatorsNSBAlexandriaVAGlobalSemiconductorSupplyChainInAnUncertainEraBCGandSIAApril7.MITREEngenuity.2021.AmericanInnovation,AmericanGrowth:AVisionfortheNationalSemiconductorTechnologyCenter.8.id.“StrengtheningTheGlobalSemiconductorSupplyChainInAnUncertainEra”9.id.“StrengtheningTheGlobalSemiconductorSupplyChainInAnUncertainEra”10.Foundriesareresponsibleforalargeshareofadvancedmanufacturingandcontrol78%globalmarketfor14nanometerorbelowtechnologynodesfor5nanometertechnologynodes11.SEMATECHreceivedpublicfundingfrom1988-1996,thenitoperatedwithoutfutherpublicsectorfunding8ssemiconductorprograms.1.Includesnationallaboratorieswithsemiconductorfabricationfacilities2.IncludesuniversitiesthatparticularinSemiconductorResearchCorporation(SRC)programsorthathostNationalNanotechnologyCoordinatedInfrastructuresiteNote:Nomajorsemiconductor-relatedfabs,universities,oremploymentinAlaska9Awell-functioningR&Decosystemsupportsinnovationbyprovidingdirection,resourcing,andcollaborationtoR&Defforts.TheU.S.retainsastrongportfolioofnationallabs,universities,andcompanies,butitsR&Decosystemcurrentlyfaceschallengesindeliveringneededinnovations.Examplesofthesechallengesinclude:icinvestmentcalmissionsandoftentakendivesareoftendistinctfromthoseofprivateationsproduction-scale.estartupstoiterateancementsImportantopportunitiesforinnovationexistsoftheossgthenomitigatetheseofoverallsemiconductorsacrossthecomputingstacktocomputingstack.WhiletheU.S.continuestooutspendotherregionsinpublicinvestmentinsemiconductorR&Dinabsoluteterms,otherregionsprovidemoregeneroussupportforR&D.12Otherregionsareexpandingtheirinvestmentsinsemiconductorstoo.Justsince2021,Japan,Singapore,SouthKorea,theEuropeanUnion(inadditiontoindividualEUmemberstateslikeSpain)haveallannouncedlegislationtosupporttheirdomesticsemiconductorcapabilities,includingthroughinvestmentsinR&D.MainlandChinatooisinvestingoverahundredbilliondollarstosupportitsdomesticsemiconductorindustry.gnitionoftheimportanceofpubliclyfundednmitigateess12.SeeAppendix:Supportforsemiconductorsinotherregionseandain3131ductorRDtorgiconductorsforPSActconstitutesthemostsignificantfederaleffortinnufacturingdedtoearliericfinancialassistanceforconstructingologiesfromlabstothemarketplaceorexpansionofpublicinvestmentinDManufacturingUSAInstitutesManufacturingUSAInstitutesamsrNSTCApublic-privatepartnershiptofosterU.S.leadershipinsemiconductorsbypromotingadvancedR&DandprototypingtostrengthenU.S.technologicalcompetitivenessandsupplychainsecurity.ingManufacturingProgram(NAPMP).ApartofNIST,NAPMPisestablishedtostrengthensemiconductoradvancedtest,assembly,andpackagingcapabilitiesintheU.S.R&Decosystem.theystemputfromcancompetitiveness.13.“History.”ManufacturingUSA,/pages/history.Accessed30June2022.14.RequestforInformation(RFI)notice“Incentives,Infrastructure,andResearchandDevelopmentNeedstoSupportaStrongDomesticSemiconductorIndustry”waspostedbytheDepartmentofCommerceonJanuary23,2022.ResponsesarepubliclyavailableonRUSAInstituteinitiativewasoriginallyestablishedin2014toboosttheU.S.manufacturingsector’sglobalcompetitivenessandfosterinnovation.13Sinceinception,16institutesacrosstheU.S.havebeencreatedtoinvigorateU.S.manufacturing,eachwithinaspecificindustrysector.InstitutesareacriticalresourceforAmericaninnovation,supportingtheinitialstagesoftheinnovationpipelineandbuildingnewandmuchneededcommunitieswherenoneexist.Theyservetoidentifypromisinghigh-impacttechnologies,thenmanagecorrespondingR&Dprojectsperformedinacademicandindustryresearchlabstofurtheradvancethesetechnologies.Uponsuccessfulcompletionoftheseprojectsthetechnologiescreatedcantransfertodevelopmentfacilitiesinindustryorotherwisefortesting,piloting,andscalingtomanufacturing.enceActauthorizedtheestablishmentofupto3ManufacturingUSAInstitutestosupportworkrelatingtosemiconductormanufacturingincludingincreasedautomation,advancedassemblyandtest,andworkforceskillstraining.FundingfortheseinstitutesisincludedaspartoftheoverallR&DfundingwithintheCHIPSandScienceAct.ThespecificprojectsdevelopedthroughthisManufacturingUSAInstitutewouldserveasakeypipelineforinnovativetechnologyinthissectorandwouldbeideallytransferredtoeithertheNSTCorNAPMPfortesting,piloting,andscaling.TheroleofthisInstituteandothersinthetechnologyinnovationprocessisclarifiedinExhibit4osteffectivelyeeratedpublicprivatesectorconsortiumwithparticipationfromtheprivateU.S.semiconductorcompaniescurrentlypartnerwithseveralexistingandwell-regardedorganizationslikeInteruniversityMicroelectronicsCenter(imec)andtheSemiconductorResearchCorporation(SRC)formanypre-competitiveR&Dneeds.TosupportU.S.technologicalandeconomiccompetitivenessmosteffectively,theNSTCandNAPMPshouldaugmentratherthanreplicatetheseorganizations.Inotherwords,whiletheNSTCandNAPMPinfrastructurecansupportearly-stageresearch,theirprimaryfocusshouldbeonmaturingandscalingthosetechnologiesthatarereadytomovebeyondearly-stageresearch.TheNSTCandNAPMPshouldbringtogetherresearchcentersacrossgovernment,academia,andindustrytoassesswhichtechnologiesofcommercialinteresttocompaniesintheU.S.mayneedbutnotbereceivingfundingfortechnologytransitionandscaling.ssedyearsfromproductionThisfuturefocusfornalingesultfrome?Advancedarchitecturesforlogic,memory,analog?3Dstackeddevices?Monolithicallyintegratedfunctions?Memory-centriccomputing?AdvancedmaterialsforbeyondComplementaryMetalOxideSemiconductor(CMOS)computingandnovelparadigms?Twodimensionalmaterials?Generalprocessdevelopments??Generalprocessdevelopments?Designinnovations?Toolimprovements?EnvironmentalSustainability?Processgaseswithalowerglobalwarmingpotential?Photolithographyandotherchemicalsthatmeetfunctionalneedswithanimprovedenvironmentalprofile,aswellasdetectionandtreatmenttechnologiesatextremelylowconcentrations?Fabprocessesthatsatisfydemandingoperationalrequirementswhileconservingnaturalresourcesnductormaterialandprocessadvancese?Advancedtestingandvalidationcapabilities?Design-for-testanddataanalysisfordesignerrorreduction?TestingautomationandintegrationofAI/MLtools?Analog,RF,andmixedsignaltesting?Heterogenousintegration?Developmentofindustrystandardsforintegration?ChipletIPdevelopmentandaccess?Integrationmethodsfornovelcomputingparadigms(photonic,quantum,etc.)?Advancedpackagingandhigh-densityinterconnects(<100μmIOpitch)?Panelandwaferlevelhigh-bandwidth,low-latencyhigh-density2.5Dand3Dstackingandassemblymethods?Hybridbonding,through-silicon-via,andadvancedinterposerdevelopment?Advancedthermalcompressionbondingfordevicelifetimeimprovements?Thermalmanagementandcrosstalk,noise,andparasiticreduction?Flexible,constrainedform-factorpackaging?ToolImprovements?Package-levelco-designtools?Superiorelectrical,thermal,mechanicalmodelinganddesigntools?Assemblyandalignmentautomation?Advancedlithographytechniques?DevelopmentofadvancedlightsourcesandEUVimprovements?Improvementsinmetallization?Advancedfunctionalmaterials?Alternateparadigmssuchasphotonicorneuromorphic?High-voltageand-powermaterials?AdvancedRFmaterials?Superiordomainspecificacceleratorsacrosswidervarietyofapplications?Mixed-signaldesigns,integrationofintelligenceandsensingcapabilities?Designforsecurity?IntegrationofAIintodesigntools,higherdesignabstraction?SuperiortoolsforanalogandRFcircuits?Toolswithenhancedcapabilityforfull-stackoptimizationandenablementofhardware-softwarecodesignationandcrosspollinationofideasmaximizemaximizetheimpactofafiniteamountofpublicinvestment.tivetocostAsanillustrativeelcapabilitiesstructureeThexternalcapabilitiesmultiprojectwafersstocommercialdesigntoolsforpre-competitivetechnologies.ybondingstrateprocessandflowsaystoprototypingfacilitiesorheymustetechnologytransitionfromlabtofab–suchasprototypingandolSemiconductorR&Dcapabilitiesaredistributedwidely–technicallyandgeographically–acrosstheU.S.ecosystem.Tostewardpublicinvestmentwell,theNSTCmustneitherspreadfundingevenlynorconcentrateinvestmentsinasingletechnologyorlocation.UndertheguidanceoftheNSTC/NAPMPsteeringcommittees,theNSTCshouldexpandandupgradethedistinctcapabilitiesandinfrastructureofalimitednumberofexistinginstitutions.latedtocostandtimeImprovingefacilitiescouldconstructionwasfinished.17ureingovationacrossthefullcomputingstackfrommaterialsductoradvancementsmechanismstocoordinatethiscollaboration.istheeconomicsofdomesticmanufacturingsectthistobecompetitiveFurthermoreTherearemanysuccessfulmodelsemployedtodayforsemiconductor-servingorganizationsthattheNSTCandNAPMPcouldlooktoforguidance.Oneexampleisimec,whichoffersmanyoftheR&Dandprototypingcapabilitiesthatarecriticalforthetechnologydevelopmentpipeline.OtherexamplesincludetheManufacturingUSAinstitutes,whichcreateregionalinnovationhubsbypartneringwithresearchinstitutionsandstandingupnewprototypinginfrastructure,andSandiaNationalLaboratory’sMic