본문

서브메뉴

Accelerating Deep Learning in the Post-Moore's Era of Computing
Accelerating Deep Learning in the Post-Moore's Era of Computing
Accelerating Deep Learning in the Post-Moore's Era of Computing

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260209102839
ISBN  
9798314843642
DDC  
004
저자명  
Hegde, Kartik.
서명/저자  
Accelerating Deep Learning in the Post-Moores Era of Computing
발행사항  
[Sl] : University of Illinois at Urbana-Champaign, 2023
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2023
형태사항  
152 p
주기사항  
Source: Dissertations Abstracts International, Volume: 86-11, Section: B.
주기사항  
Advisor: Fletcher, Christopher W.
학위논문주기  
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2023.
초록/해제  
요약Rapid growth in the availability of cheap and fast computing power-made possible by Moore's scaling-has been a major tailwind behind the rise of modern deep learning. However, over the last decade, the performance deficit between compute requirements of deep learning workloads and the peak compute capability offered by modern processors has been rapidly increasing due to two opposing trends. On the one hand, improving the accuracy of deep learning models requires larger models and datasets, which in turn requires more compute resources. On the other hand, Moore's scaling has been slowing down, leading to a slowdown in the rate of increase in peak compute throughput. While specialization has been seen as a way to address the increasing performance deficit, there is mounting evidence that specialization has only afforded us a "one-time" boost in performance. To make matters worse, the increasing gap between logic and memory, often called the memory wall, makes it harder to utilize all the compute on-chip, let alone add more. Therefore, the key question facing deep learning architects is, how do we continue to accelerate deep learning workloads in the post-Moore's era of computing?This thesis presents a set of techniques to accelerate deep learning workloads without scaling the number of transistors on-chip. At a high level, we classify them into two thrusts. First, techniques that reduce the total compute required for deep learning workloads without sacrificing accuracy and utilize that to improve performance. Second, techniques that increase the compute-per-transistor by improving the resource utilization of deep learning accelerators. Notably, we show that the proposed techniques are not "zero-sum", i.e., the gained performance comes at negligible area/power overheads. We begin by describing a baseline deep learning accelerator (DLA) tailored for deep learning workloads and represents state-of-the-art deep learning accelerators. We then describe concrete ways to evolve the baseline architecture to incorporate our proposed techniques. Finally, we provide a detailed analysis of the area overheads of the proposed techniques and demonstrate that they are minimal while the performance gains are significant.Overall, this thesis throws light on different directions that computer architects can take to continue to accelerate deep learning workloads without scaling the number of transistors on-chip. We believe that expanding the scope of specialization beyond hardware to include other layers of the deep learning stack and careful co-design will enable us to continue to accelerate deep learning workloads in the post-Moore's era of computing.
일반주제명  
Computer science
일반주제명  
Engineering
일반주제명  
Information technology
키워드  
Deep learning
키워드  
Computer architecture
키워드  
Moore's law
키워드  
Specialized accelerators
키워드  
Deep neural networks
키워드  
Programmable accelerators
기타저자  
University of Illinois at Urbana-Champaign Computer Science
기본자료저록  
Dissertations Abstracts International. 86-11B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260203s2023        us                              c    eng  d
■001000017365854
■00520260209102839
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798314843642
■035    ▼a(MiAaPQ)AAI32063794
■035    ▼a(MiAaPQ)httphdlhandlenet2142120125
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a004
■1001  ▼aHegde,  Kartik.
■24510▼aAccelerating  Deep  Learning  in  the  Post-Moore's  Era  of  Computing
■260    ▼a[Sl]▼bUniversity  of  Illinois  at  Urbana-Champaign▼c2023
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2023
■300    ▼a152  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  86-11,  Section:  B.
■500    ▼aAdvisor:  Fletcher,  Christopher  W.
■5021  ▼aThesis  (Ph.D.)--University  of  Illinois  at  Urbana-Champaign,  2023.
■520    ▼aRapid  growth  in  the  availability  of  cheap  and  fast  computing  power-made  possible  by  Moore's  scaling-has  been  a  major  tailwind  behind  the  rise  of  modern  deep  learning.  However,  over  the  last  decade,  the  performance  deficit  between  compute  requirements  of  deep  learning  workloads  and  the  peak  compute  capability  offered  by  modern  processors  has  been  rapidly  increasing  due  to  two  opposing  trends.  On  the  one  hand,  improving  the  accuracy  of  deep  learning  models  requires  larger  models  and  datasets,  which  in  turn  requires  more  compute  resources.  On  the  other  hand,  Moore's  scaling  has  been  slowing  down,  leading  to  a  slowdown  in  the  rate  of  increase  in  peak  compute  throughput.  While  specialization  has  been  seen  as  a  way  to  address  the  increasing  performance  deficit,  there  is  mounting  evidence  that  specialization  has  only  afforded  us  a  "one-time"  boost  in  performance.  To  make  matters  worse,  the  increasing  gap  between  logic  and  memory,  often  called  the  memory  wall,  makes  it  harder  to  utilize  all  the  compute  on-chip,  let  alone  add  more.  Therefore,  the  key  question  facing  deep  learning  architects  is,  how  do  we  continue  to  accelerate  deep  learning  workloads  in  the  post-Moore's  era  of  computing?This  thesis  presents  a  set  of  techniques  to  accelerate  deep  learning  workloads  without  scaling  the  number  of  transistors  on-chip.  At  a  high  level,  we  classify  them  into  two  thrusts.  First,  techniques  that  reduce  the  total  compute  required  for  deep  learning  workloads  without  sacrificing  accuracy  and  utilize  that  to  improve  performance.  Second,  techniques  that  increase  the  compute-per-transistor  by  improving  the  resource  utilization  of  deep  learning  accelerators.  Notably,  we  show  that  the  proposed  techniques  are  not  "zero-sum",  i.e.,  the  gained  performance  comes  at  negligible  area/power  overheads.  We  begin  by  describing  a  baseline  deep  learning  accelerator  (DLA)  tailored  for  deep  learning  workloads  and  represents  state-of-the-art  deep  learning  accelerators.  We  then  describe  concrete  ways  to  evolve  the  baseline  architecture  to  incorporate  our  proposed  techniques.  Finally,  we  provide  a  detailed  analysis  of  the  area  overheads  of  the  proposed  techniques  and  demonstrate  that  they  are  minimal  while  the  performance  gains  are  significant.Overall,  this  thesis  throws  light  on  different  directions  that  computer  architects  can  take  to  continue  to  accelerate  deep  learning  workloads  without  scaling  the  number  of  transistors  on-chip.  We  believe  that  expanding  the  scope  of  specialization  beyond  hardware  to  include  other  layers  of  the  deep  learning  stack  and  careful  co-design  will  enable  us  to  continue  to  accelerate  deep  learning  workloads  in  the  post-Moore's  era  of  computing.
■590    ▼aSchool  code:  0090.
■650  4▼aComputer  science
■650  4▼aEngineering
■650  4▼aInformation  technology
■653    ▼aDeep  learning
■653    ▼aComputer  architecture
■653    ▼aMoore's  law
■653    ▼aSpecialized  accelerators
■653    ▼aDeep  neural  networks
■653    ▼aProgrammable  accelerators
■690    ▼a0984
■690    ▼a0489
■690    ▼a0800
■690    ▼a0537
■71020▼aUniversity  of  Illinois  at  Urbana-Champaign▼bComputer  Science.
■7730  ▼tDissertations  Abstracts  International▼g86-11B.
■790    ▼a0090
■791    ▼aPh.D.
■792    ▼a2023
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17365854▼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
    TF17069 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.