본문

서브메뉴

Thermal Management and Transistor Modeling for Three-Dimensional Integrated Circuits
Thermal Management and Transistor Modeling for Three-Dimensional Integrated Circuits
Thermal Management and Transistor Modeling for Three-Dimensional Integrated Circuits

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260202104739
ISBN  
9798290651866
DDC  
621
저자명  
Köroğlu, Çağil.
서명/저자  
Thermal Management and Transistor Modeling for Three-Dimensional Integrated Circuits
발행사항  
[Sl] : Stanford University, 2024
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2024
형태사항  
138 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-04, Section: B.
주기사항  
Advisor: Pop, Eric.
학위논문주기  
Thesis (Ph.D.)--Stanford University, 2024.
초록/해제  
요약In engineering, one usually seeks compromises to meet specifications for a certain application. A relatively unusual thing about integrated circuits (IC) before 2000s is that there was such a thing as "free lunch." As we made transistors smaller, they got cheaper, faster, more energy efficient, and more of them could be packed onto an IC to drive an explosive increase in functionality and accessibility for decades, with essentially no tradeoffs. There were humps along the way, calling for technological advancements to make this possible, but life was relatively easy. As we reach physical limits in transistor size and temperature that hinder conventional two-dimensional scaling, this is no longer the case.To meet the ever-increasing demand for computation around the world, and curb the associated rise in energy consumption, new scaling paradigms are needed to cram more devices onto a semiconductor package and increase energy efficiency. It seems clear that the path forward is in going three-dimensional (3D), stacking logic and memory elements vertically to dramatically increase transistor density and connectivity. However, going vertical does come with significant engineering challenges that need to be ironed out to enable high-volume production of dense, monolithic 3D ICs, allowing scalable advancements in computing. This thesis presents thermal and electrical models to explore and address certain aspects of these issues.I begin with a historical perspective on the exigent challenges that exist in computing today. This is followed by a discussion of the key barriers to broader uptake of 3D ICs that I have focused on during my PhD work, along with proposed solutions. 3D ICs have special cooling needs due to the difficulty of removing heat from the upper layers of the 3D stack, as well as increased transistor count (and hence power) per unit area. Moreover, monolithic 3D integration calls for transistors that can be fabricated at low temperature on noncrystalline materials and/or non-planar features, for which two-dimensional (2D) materials are promising candidates.Next, I explore passive conduction-based cooling solutions for 3D ICs that utilize high-thermal conductivity insulators as the interlayer dielectric. I show that isotropic insulators like AlN can essentially eliminate the inter-tier temperature differences that result due to the thermal resistance between the different layers in the 3D stack, which is especially useful for 3D ICs based on a logic-on-logic architecture. On the other hand, highly anisotropic insulators such as hexagonal BN (hBN) can be useful in a memory-on-logic architecture, cooling the logic layer(s) while simultaneously providing a degree of thermal isolation between logic and memory to keep the memory layer(s) cool.I then discuss the unique thermal challenges of 2D semiconductor transistors, which are promising candidates for back-end-of-line (BEOL) compatible transistors for 3D ICs. I show that the interfaces and the contact electrodes play an important role in cooling these devices, and that keeping these devices cool is largely a problem of optimizing the interfaces of the semiconductor. I also present fast analytical thermal models of these devices that can be packaged into electrical compact models to accurately capture self-heating effects in circuit simulations.Finally, I quantify fringe currents in transistors with an unpatterned channel, which can prevent the accurate evaluation of materials and fabrication processes for transistors. Such material discovery is important for progress towards monolithic 3D ICs, given their need for new transistor technologies. I show that contact resistance and velocity saturation can more than double fringe currents in such devices, causing mobility to be overestimated by up to 70% in typical devices. I conclude with a summary of my contributions, and a perspective on future progress required to enable widespread adoption of 3D ICs, as well as monolithic 3D integration. I provide a list of my major contributions to other studies in an appendix at the end of this thesis.
일반주제명  
Heat transfer
일반주제명  
Integrated circuits
일반주제명  
Semiconductor research
일반주제명  
Diamonds
일반주제명  
Cooling
일반주제명  
Conductivity
일반주제명  
Microscopy
일반주제명  
Transistors
일반주제명  
Indium tin oxides
일반주제명  
Heat conductivity
일반주제명  
Thin films
일반주제명  
Geometry
일반주제명  
Thermodynamics
일반주제명  
Electrical engineering
키워드  
High-thermal conductivity
키워드  
Thermal management
키워드  
Integrated circuits
기타저자  
Stanford University.
기본자료저록  
Dissertations Abstracts International. 87-04B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2024        us                              c    eng  d
■001000017358698
■00520260202104739
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798290651866
■035    ▼a(MiAaPQ)AAI32149685
■035    ▼a(MiAaPQ)Stanfordmk057nv5270
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a621
■1001  ▼aKöroğlu,  Çağil.
■24510▼aThermal  Management  and  Transistor  Modeling  for  Three-Dimensional  Integrated  Circuits
■260    ▼a[Sl]▼bStanford  University▼c2024
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2024
■300    ▼a138  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-04,  Section:  B.
■500    ▼aAdvisor:  Pop,  Eric.
■5021  ▼aThesis  (Ph.D.)--Stanford  University,  2024.
■520    ▼aIn  engineering,  one  usually  seeks  compromises  to  meet  specifications  for  a  certain  application.  A  relatively  unusual  thing  about  integrated  circuits  (IC)  before  2000s  is  that  there  was  such  a  thing  as  "free  lunch."  As  we  made  transistors  smaller,  they  got  cheaper,  faster,  more  energy  efficient,  and  more  of  them  could  be  packed  onto  an  IC  to  drive  an  explosive  increase  in  functionality  and  accessibility  for  decades,  with  essentially  no  tradeoffs.  There  were  humps  along  the  way,  calling  for  technological  advancements  to  make  this  possible,  but  life  was  relatively  easy.  As  we  reach  physical  limits  in  transistor  size  and  temperature  that  hinder  conventional  two-dimensional  scaling,  this  is  no  longer  the  case.To  meet  the  ever-increasing  demand  for  computation  around  the  world,  and  curb  the  associated  rise  in  energy  consumption,  new  scaling  paradigms  are  needed  to  cram  more  devices  onto  a  semiconductor  package  and  increase  energy  efficiency.  It  seems  clear  that  the  path  forward  is  in  going  three-dimensional  (3D),  stacking  logic  and  memory  elements  vertically  to  dramatically  increase  transistor  density  and  connectivity.  However,  going  vertical  does  come  with  significant  engineering  challenges  that  need  to  be  ironed  out  to  enable  high-volume  production  of  dense,  monolithic  3D  ICs,  allowing  scalable  advancements  in  computing.  This  thesis  presents  thermal  and  electrical  models  to  explore  and  address  certain  aspects  of  these  issues.I  begin  with  a  historical  perspective  on  the  exigent  challenges  that  exist  in  computing  today.  This  is  followed  by  a  discussion  of  the  key  barriers  to  broader  uptake  of  3D  ICs  that  I  have  focused  on  during  my  PhD  work,  along  with  proposed  solutions.  3D  ICs  have  special  cooling  needs  due  to  the  difficulty  of  removing  heat  from  the  upper  layers  of  the  3D  stack,  as  well  as  increased  transistor  count  (and  hence  power)  per  unit  area.  Moreover,  monolithic  3D  integration  calls  for  transistors  that  can  be  fabricated  at  low  temperature  on  noncrystalline  materials  and/or  non-planar  features,  for  which  two-dimensional  (2D)  materials  are  promising  candidates.Next,  I  explore  passive  conduction-based  cooling  solutions  for  3D  ICs  that  utilize  high-thermal  conductivity  insulators  as  the  interlayer  dielectric.  I  show  that  isotropic  insulators  like  AlN  can  essentially  eliminate  the  inter-tier  temperature  differences  that  result  due  to  the  thermal  resistance  between  the  different  layers  in  the  3D  stack,  which  is  especially  useful  for  3D  ICs  based  on  a  logic-on-logic  architecture.  On  the  other  hand,  highly  anisotropic  insulators  such  as  hexagonal  BN  (hBN)  can  be  useful  in  a  memory-on-logic  architecture,  cooling  the  logic  layer(s)  while  simultaneously  providing  a  degree  of  thermal  isolation  between  logic  and  memory  to  keep  the  memory  layer(s)  cool.I  then  discuss  the  unique  thermal  challenges  of  2D  semiconductor  transistors,  which  are  promising  candidates  for  back-end-of-line  (BEOL)  compatible  transistors  for  3D  ICs.  I  show  that  the  interfaces  and  the  contact  electrodes  play  an  important  role  in  cooling  these  devices,  and  that  keeping  these  devices  cool  is  largely  a  problem  of  optimizing  the  interfaces  of  the  semiconductor.  I  also  present  fast  analytical  thermal  models  of  these  devices  that  can  be  packaged  into  electrical  compact  models  to  accurately  capture  self-heating  effects  in  circuit  simulations.Finally,  I  quantify  fringe  currents  in  transistors  with  an  unpatterned  channel,  which  can  prevent  the  accurate  evaluation  of  materials  and  fabrication  processes  for  transistors.  Such  material  discovery  is  important  for  progress  towards  monolithic  3D  ICs,  given  their  need  for  new  transistor  technologies.  I  show  that  contact  resistance  and  velocity  saturation  can  more  than  double  fringe  currents  in  such  devices,  causing  mobility  to  be  overestimated  by  up  to  70%  in  typical  devices.  I  conclude  with  a  summary  of  my  contributions,  and  a  perspective  on  future  progress  required  to  enable  widespread  adoption  of  3D  ICs,  as  well  as  monolithic  3D  integration.  I  provide  a  list  of  my  major  contributions  to  other  studies  in  an  appendix  at  the  end  of  this  thesis.
■590    ▼aSchool  code:  0212.
■650  4▼aHeat  transfer
■650  4▼aIntegrated  circuits
■650  4▼aSemiconductor  research
■650  4▼aDiamonds
■650  4▼aCooling
■650  4▼aConductivity
■650  4▼aMicroscopy
■650  4▼aTransistors
■650  4▼aIndium  tin  oxides
■650  4▼aHeat  conductivity
■650  4▼aThin  films
■650  4▼aGeometry
■650  4▼aThermodynamics
■650  4▼aElectrical  engineering
■653    ▼aHigh-thermal  conductivity
■653    ▼aThermal  management
■653    ▼aIntegrated  circuits
■690    ▼a0544
■690    ▼a0348
■71020▼aStanford  University.
■7730  ▼tDissertations  Abstracts  International▼g87-04B.
■790    ▼a0212
■791    ▼aPh.D.
■792    ▼a2024
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17358698▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

미리보기

내보내기

chatGPT토론

Ai 추천 관련 도서


    신착도서 더보기
    최근 3년간 통계입니다.

    소장정보

    • 예약
    • 소재불명신고
    • 나의폴더
    • 우선정리요청
    • 비도서대출신청
    • 야간 도서대출신청
    소장자료
    등록번호 청구기호 소장처 대출가능여부 대출정보
    TF19243 전자도서 대출가능 마이폴더 부재도서신고 비도서대출신청 야간 도서대출신청

    * 대출중인 자료에 한하여 예약이 가능합니다. 예약을 원하시면 예약버튼을 클릭하십시오.

    해당 도서를 다른 이용자가 함께 대출한 도서

    관련 인기도서

    로그인 후 이용 가능합니다.