본문

서브메뉴

Enabling Full-Stack DNN Accelerator Design and Evaluation on Synthesizable Hardware
Enabling Full-Stack DNN Accelerator Design and Evaluation on Synthesizable Hardware
Enabling Full-Stack DNN Accelerator Design and Evaluation on Synthesizable Hardware

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20250211152726
ISBN  
9798384454854
DDC  
621.3
저자명  
Genc, Hasan Nazim.
서명/저자  
Enabling Full-Stack DNN Accelerator Design and Evaluation on Synthesizable Hardware
발행사항  
[Sl] : University of California, Berkeley, 2024
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2024
형태사항  
104 p
주기사항  
Source: Dissertations Abstracts International, Volume: 86-03, Section: B.
주기사항  
Advisor: Asanovic, Krste.
학위논문주기  
Thesis (Ph.D.)--University of California, Berkeley, 2024.
초록/해제  
요약The growing diversity of computationally demanding DNN workloads, together with the long-running decline of technology-scaling trends, has motivated the design of a great many diverse specialized hardware accelerators. While these accelerators provide significant improvements to performance and energy consumption - making the modern wave of AI innovation possible - they introduce significant challenges to computer architects, programmers, and hardware designers, due to the difficulty of (i) exploring the very broad design space they represent, (ii) translating such designs rapidly to high-quality RTL and software libraries, and (iii) evaluating such designs in realistic full-system contexts early on in the design process.Prior work has attempted to address these difficulties by proposing new accelerator design frameworks which allow users to change only a few settings in a config file, or a few lines of a domain-specific language, from which they can rapidly generate new synthesizable hardware, or new high-level models that can guide architectural decisions. Many of these frameworks also attempt to make accelerator design more principled by separating the different concerns which go into accelerator design, so that each can be explored individually, and in relation to other design choices.However, prior accelerator generators and design frameworks often lack the ability to provide users visibility into the impact that the full system and software stack have upon DNN accelerator performance, such as the potential for outer caches, virtual address translation mechanisms, or host CPUs to bottleneck performance if they have not been carefully tuned along with the accelerator's functional units or spatial arrays. Prior accelerator design frameworks which separate out different design concerns are also are not capable of generating high-quality RTL for both dense and sparse accelerator ASICs, which limits their ability to cover various modern workloads which sparsify DNN layers to improve performance or energy efficiency.This thesis presents two projects, Gemmini and Stellar, which address these difficulties. Gemmini is a DNN accelerator evaluation framework which, while generating efficient spatial arrays and accelerators, is primarily intended to help users evaluate the impact of SoC components outside of the accelerator itself, such as external caches or virtual address translation mechanisms, upon overall DNN accelerator performance. Stellar is another framework which provides abstractions that help users to design and explore different components of both dense and sparse accelerators, while separating out the different concerns that go into accelerator design, such as an accelerator's functionality, its dataflow, the sparse/dense data formats it supports, its load-balancing strategies, and the private memory buffers it is equipped with. Gemmini-generated dense DNN accelerators achieve 87% the performance of prior state-of-the-art accelerators such as NVDLA on image classification networks such as ResNet50, and enable insights into how minor changes to system components such as TLBs can improve end-to-end DNN performance by up to 15%. Stellar-generated accelerators achieve up to 92% the performance of hand-written accelerators, with less than 15% area overhead and power overheads on various DNN layers as low as 7%.
일반주제명  
Computer engineering
일반주제명  
Computer science
일반주제명  
Information technology
키워드  
Computer architecture
키워드  
Deep neural networks accelerators
키워드  
Hardware design methodology
키워드  
Spatial accelerators
키워드  
Full-stack evaluation
기타저자  
University of California, Berkeley Computer Science
기본자료저록  
Dissertations Abstracts International. 86-03B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008250123s2024        us                              c    eng  d
■001000017163572
■00520250211152726
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798384454854
■035    ▼a(MiAaPQ)AAI31490198
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a621.3
■1001  ▼aGenc,  Hasan  Nazim.
■24510▼aEnabling  Full-Stack  DNN  Accelerator  Design  and  Evaluation  on  Synthesizable  Hardware
■260    ▼a[Sl]▼bUniversity  of  California,  Berkeley▼c2024
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2024
■300    ▼a104  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  86-03,  Section:  B.
■500    ▼aAdvisor:  Asanovic,  Krste.
■5021  ▼aThesis  (Ph.D.)--University  of  California,  Berkeley,  2024.
■520    ▼aThe  growing  diversity  of  computationally  demanding  DNN  workloads,  together  with  the  long-running  decline  of  technology-scaling  trends,  has  motivated  the  design  of  a  great  many  diverse  specialized  hardware  accelerators.  While  these  accelerators  provide  significant  improvements  to  performance  and  energy  consumption  -  making  the  modern  wave  of  AI  innovation  possible  -  they  introduce  significant  challenges  to  computer  architects,  programmers,  and  hardware  designers,  due  to  the  difficulty  of  (i)  exploring  the  very  broad  design  space  they  represent,  (ii)  translating  such  designs  rapidly  to  high-quality  RTL  and  software  libraries,  and  (iii)  evaluating  such  designs  in  realistic  full-system  contexts  early  on  in  the  design  process.Prior  work  has  attempted  to  address  these  difficulties  by  proposing  new  accelerator  design  frameworks  which  allow  users  to  change  only  a  few  settings  in  a  config  file,  or  a  few  lines  of  a  domain-specific  language,  from  which  they  can  rapidly  generate  new  synthesizable  hardware,  or  new  high-level  models  that  can  guide  architectural  decisions.  Many  of  these  frameworks  also  attempt  to  make  accelerator  design  more  principled  by  separating  the  different  concerns  which  go  into  accelerator  design,  so  that  each  can  be  explored  individually,  and  in  relation  to  other  design  choices.However,  prior  accelerator  generators  and  design  frameworks  often  lack  the  ability  to  provide  users  visibility  into  the  impact  that  the  full  system  and  software  stack  have  upon  DNN  accelerator  performance,  such  as  the  potential  for  outer  caches,  virtual  address  translation  mechanisms,  or  host  CPUs  to  bottleneck  performance  if  they  have  not  been  carefully  tuned  along  with  the  accelerator's  functional  units  or  spatial  arrays.  Prior  accelerator  design  frameworks  which  separate  out  different  design  concerns  are  also  are  not  capable  of  generating  high-quality  RTL  for  both  dense  and  sparse  accelerator  ASICs,  which  limits  their  ability  to  cover  various  modern  workloads  which  sparsify  DNN  layers  to  improve  performance  or  energy  efficiency.This  thesis  presents  two  projects,  Gemmini  and  Stellar,  which  address  these  difficulties.  Gemmini  is  a  DNN  accelerator  evaluation  framework  which,  while  generating  efficient  spatial  arrays  and  accelerators,  is  primarily  intended  to  help  users  evaluate  the  impact  of  SoC  components  outside  of  the  accelerator  itself,  such  as  external  caches  or  virtual  address  translation  mechanisms,  upon  overall  DNN  accelerator  performance.  Stellar  is  another  framework  which  provides  abstractions  that  help  users  to  design  and  explore  different  components  of  both  dense  and  sparse  accelerators,  while  separating  out  the  different  concerns  that  go  into  accelerator  design,  such  as  an  accelerator's  functionality,  its  dataflow,  the  sparse/dense  data  formats  it  supports,  its  load-balancing  strategies,  and  the  private  memory  buffers  it  is  equipped  with.  Gemmini-generated  dense  DNN  accelerators  achieve  87%  the  performance  of  prior  state-of-the-art  accelerators  such  as  NVDLA  on  image  classification  networks  such  as  ResNet50,  and  enable  insights  into  how  minor  changes  to  system  components  such  as  TLBs  can  improve  end-to-end  DNN  performance  by  up  to  15%.  Stellar-generated  accelerators  achieve  up  to  92%  the  performance  of  hand-written  accelerators,  with  less  than  15%  area  overhead  and  power  overheads  on  various  DNN  layers  as  low  as  7%.
■590    ▼aSchool  code:  0028.
■650  4▼aComputer  engineering
■650  4▼aComputer  science
■650  4▼aInformation  technology
■653    ▼aComputer  architecture
■653    ▼aDeep  neural  networks  accelerators
■653    ▼aHardware  design  methodology
■653    ▼aSpatial  accelerators
■653    ▼aFull-stack  evaluation
■690    ▼a0464
■690    ▼a0984
■690    ▼a0489
■71020▼aUniversity  of  California,  Berkeley▼bComputer  Science.
■7730  ▼tDissertations  Abstracts  International▼g86-03B.
■790    ▼a0028
■791    ▼aPh.D.
■792    ▼a2024
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17163572▼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
    TF09614 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.