본문

서브메뉴

Domain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data
Domain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data
Domain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260202105232
ISBN  
9798297610064
DDC  
621.3
저자명  
McCrabb, Andrew.
서명/저자  
Domain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data
발행사항  
[Sl] : University of Michigan, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
214 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-04, Section: B.
주기사항  
Advisor: Bertacco, Valeria M.
학위논문주기  
Thesis (Ph.D.)--University of Michigan, 2025.
초록/해제  
요약Graph-based processing enables many applications in logistics, e-commerce, social media, and more. However, graph workloads are slow: they are bottlenecked not by compute power, but by inefficient data access. As useful graphs get larger and graph-based algorithms become more complex, adding more powerful compute units like Graphics Processing Units (GPUs) or Tensor Processing Units (TPUs) cannot keep up with the increasing size and complexity of these workloads.To address these challenges, we first introduce a novel taxonomy of graph-based algorithms: those that treat graphs (1) as data frameworks, (2) as algorithmic frameworks, or (3) as both. Each category has overlapping needs: higher memory bandwidth, better data organization, and greater thread-level parallelism. Next, we demonstrate that custom processing-in-memory (PIM) hardware accelerators are effective and energy-efficient solutions to the compute and memory bottlenecks of graph-based applications.Specifically, we propose and evaluate three custom PIM accelerators, DREDGE (for graph-as-data-framework applications), ACRE (for graph-as-algorithmic-framework applications), and GLEAM (for graph-as-both applications), each targeting one of the three categories of graph applications. DREDGE targets dynamic graph applications by introducing a novel partitioning technique and dedicated hardware support to continuously improve data organization in memory. ACRE accelerates the training of tree-based machine learning models in a way that allows users to better understand the models' reasoning. GLEAM targets graph neural networks, the primary machine learning models for graph-based data, accelerating the node aggregation operations that bottleneck training and inference operations. These three designs offer a 2.5-14x speedup for their respective applications, and they save 77-93% of total system energy over their respective baselines. Each design fits within the logic area of modern 3D-stacked memory: 0.3-13% of the available logic space. Finally, we present two benchmark suites, DyGraph and BeXAI, making them publicly available to support future research in dynamic graphs processing and explainable machine learning acceleration. Together, these contributions enable efficient and scalable graph computing to handle the demands of tomorrow's graph workloads.
일반주제명  
Computer engineering
일반주제명  
Computer science
일반주제명  
Information technology
키워드  
Memory
키워드  
Graph processing
키워드  
Hardware accelerators
키워드  
Graphics Processing Units
키워드  
Tensor Processing Units
기타저자  
University of Michigan Computer Science & Engineering
기본자료저록  
Dissertations Abstracts International. 87-04B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017359894
■00520260202105232
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798297610064
■035    ▼a(MiAaPQ)AAI32271908
■035    ▼a(MiAaPQ)umichrackham006396
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a621.3
■1001  ▼aMcCrabb,  Andrew.
■24510▼aDomain-Specific  Benchmarks  and  Architectures  for  Applications  Using  Graph-Based  Data
■260    ▼a[Sl]▼bUniversity  of  Michigan▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a214  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-04,  Section:  B.
■500    ▼aAdvisor:  Bertacco,  Valeria  M.
■5021  ▼aThesis  (Ph.D.)--University  of  Michigan,  2025.
■520    ▼aGraph-based  processing  enables  many  applications  in  logistics,  e-commerce,  social  media,  and  more.  However,  graph  workloads  are  slow:  they  are  bottlenecked  not  by  compute  power,  but  by  inefficient  data  access.  As  useful  graphs  get  larger  and  graph-based  algorithms  become  more  complex,  adding  more  powerful  compute  units  like  Graphics  Processing  Units  (GPUs)  or  Tensor  Processing  Units  (TPUs)  cannot  keep  up  with  the  increasing  size  and  complexity  of  these  workloads.To  address  these  challenges,  we  first  introduce  a  novel  taxonomy  of  graph-based  algorithms:  those  that  treat  graphs  (1)  as  data  frameworks,  (2)  as  algorithmic  frameworks,  or  (3)  as  both.  Each  category  has  overlapping  needs:  higher  memory  bandwidth,  better  data  organization,  and  greater  thread-level  parallelism.  Next,  we  demonstrate  that  custom  processing-in-memory  (PIM)  hardware  accelerators  are  effective  and  energy-efficient  solutions  to  the  compute  and  memory  bottlenecks  of  graph-based  applications.Specifically,  we  propose  and  evaluate  three  custom  PIM  accelerators,  DREDGE  (for  graph-as-data-framework  applications),  ACRE  (for  graph-as-algorithmic-framework  applications),  and  GLEAM  (for  graph-as-both  applications),  each  targeting  one  of  the  three  categories  of  graph  applications.  DREDGE  targets  dynamic  graph  applications  by  introducing  a  novel  partitioning  technique  and  dedicated  hardware  support  to  continuously  improve  data  organization  in  memory.  ACRE  accelerates  the  training  of  tree-based  machine  learning  models  in  a  way  that  allows  users  to  better  understand  the  models'  reasoning.  GLEAM  targets  graph  neural  networks,  the  primary  machine  learning  models  for  graph-based  data,  accelerating  the  node  aggregation  operations  that  bottleneck  training  and  inference  operations.  These  three  designs  offer  a  2.5-14x  speedup  for  their  respective  applications,  and  they  save  77-93%  of  total  system  energy  over  their  respective  baselines.  Each  design  fits  within  the  logic  area  of  modern  3D-stacked  memory:  0.3-13%  of  the  available  logic  space.  Finally,  we  present  two  benchmark  suites,  DyGraph  and  BeXAI,  making  them  publicly  available  to  support  future  research  in  dynamic  graphs  processing  and  explainable  machine  learning  acceleration.  Together,  these  contributions  enable  efficient  and  scalable  graph  computing  to  handle  the  demands  of  tomorrow's  graph  workloads.
■590    ▼aSchool  code:  0127.
■650  4▼aComputer  engineering
■650  4▼aComputer  science
■650  4▼aInformation  technology
■653    ▼aMemory
■653    ▼aGraph  processing
■653    ▼aHardware  accelerators
■653    ▼aGraphics  Processing  Units
■653    ▼aTensor  Processing  Units
■690    ▼a0464
■690    ▼a0984
■690    ▼a0489
■71020▼aUniversity  of  Michigan▼bComputer  Science  &  Engineering.
■7730  ▼tDissertations  Abstracts  International▼g87-04B.
■790    ▼a0127
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17359894▼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
    TF15673 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.