본문

서브메뉴

Efficiently Scaling Machine Learning Systems Across Heterogeneous Resources
Efficiently Scaling Machine Learning Systems Across Heterogeneous Resources
Efficiently Scaling Machine Learning Systems Across Heterogeneous Resources

Detailed Information

자료유형  
 학위논문 서양
최종처리일시  
20260202105245
ISBN  
9798291569672
DDC  
621.3
저자명  
Jin, Shuowei.
서명/저자  
Efficiently Scaling Machine Learning Systems Across Heterogeneous Resources
발행사항  
[Sl] : University of Michigan, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
116 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-03, Section: B.
주기사항  
Advisor: Mao, Z. Morley.
학위논문주기  
Thesis (Ph.D.)--University of Michigan, 2025.
초록/해제  
요약The rapid advancement of machine learning (ML), particularly deep learning and Large Language Models (LLMs), has revolutionized numerous domains. This progress is driven by scaling laws, indicating that larger models trained on more data with more compute yield better performance. However, this scaling introduces significant challenges, especially during inference, due to massive model sizes and computational demands. Furthermore, emerging inference-time scaling laws, where allocating more compute during inference boosts performance, add another layer of complexity. Efficiently scaling ML inference systems to optimize critical metrics like latency and throughput is further complicated by the inherent heterogeneity of computing resources, spanning from resource-constrained edge devices to diverse hardware within datacenters. Naive scaling approaches often lead to suboptimal performance and resource utilization in these heterogeneous environments.This dissertation argues that efficiently scaling ML systems across heterogeneous resources requires a holistic approach. We adopt Algorithm-System Co-design as the central methodology, simultaneously optimizing ML algorithms and the underlying system components (hardware, software, schedulers) by considering their intricate interactions. This co-design philosophy allows us to overcome the limitations of optimizing algorithms and systems in isolation, unlocking significant efficiency gains.We demonstrate the efficacy of this approach through three core contributions. First, OASIS, a system for collaborative neural-enhanced mobile video streaming. It addresses the challenge of running demanding super-resolution models on resource-constrained mobile devices with varying network conditions. OASIS co-designs adaptive bitrate/model selection algorithms with a cross-device collaborative inference mechanism, enabling multiple nearby devices to efficiently share computation resources, significantly improving video Quality of Experience (QoE) and reducing energy consumption. Second, Cake, a system optimizing Large Language Model (LLM) inference latency, specifically targeting the Key-Value (KV) cache generation bottleneck during long-context prefill. It tackles the compute vs. I/O trade-off in heterogeneous memory systems by co-designing a bidirectional parallel scheduling strategy that dynamically balances computing the KV cache on GPUs and loading it from storage, substantially reducing Time-To-First-Token (TTFT). Third, Plato, a semantic-aware parallel decoding framework for LLMs that addresses the inefficiency of autoregressive decoding while maintaining answer quality. Plato co-designs algorithmic components (organizing sub-problems into dependency graphs based on logical relationships) with system optimizations (pipelined execution and KV cache reuse). Our evaluations show that Plato improves throughput by up to 68% over autoregressive decoding while maintaining a 90% quality net-win rate compared to other parallel decoding methods like Skeleton-of-Thought.Collectively, these contributions illustrate that principled algorithm-system co-design is a powerful and versatile methodology for building efficient ML inference systems that can effectively scale across diverse and heterogeneous resources, optimizing for latency and throughput from the edge to the datacenter.
일반주제명  
Computer engineering
일반주제명  
Computer science
키워드  
Machine learning systems
키워드  
Large Language Models
키워드  
Energy consumption
기타저자  
University of Michigan Computer Science & Engineering
기본자료저록  
Dissertations Abstracts International. 87-03B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017359981
■00520260202105245
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798291569672
■035    ▼a(MiAaPQ)AAI32272042
■035    ▼a(MiAaPQ)umichrackham006341
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a621.3
■1001  ▼aJin,  Shuowei.
■24510▼aEfficiently  Scaling  Machine  Learning  Systems  Across  Heterogeneous  Resources
■260    ▼a[Sl]▼bUniversity  of  Michigan▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a116  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-03,  Section:  B.
■500    ▼aAdvisor:  Mao,  Z.  Morley.
■5021  ▼aThesis  (Ph.D.)--University  of  Michigan,  2025.
■520    ▼aThe  rapid  advancement  of  machine  learning  (ML),  particularly  deep  learning  and  Large  Language  Models  (LLMs),  has  revolutionized  numerous  domains.  This  progress  is  driven  by  scaling  laws,  indicating  that  larger  models  trained  on  more  data  with  more  compute  yield  better  performance.  However,  this  scaling  introduces  significant  challenges,  especially  during  inference,  due  to  massive  model  sizes  and  computational  demands.  Furthermore,  emerging  inference-time  scaling  laws,  where  allocating  more  compute  during  inference  boosts  performance,  add  another  layer  of  complexity.  Efficiently  scaling  ML  inference  systems  to  optimize  critical  metrics  like  latency  and  throughput  is  further  complicated  by  the  inherent  heterogeneity  of  computing  resources,  spanning  from  resource-constrained  edge  devices  to  diverse  hardware  within  datacenters.  Naive  scaling  approaches  often  lead  to  suboptimal  performance  and  resource  utilization  in  these  heterogeneous  environments.This  dissertation  argues  that  efficiently  scaling  ML  systems  across  heterogeneous  resources  requires  a  holistic  approach.  We  adopt  Algorithm-System  Co-design  as  the  central  methodology,  simultaneously  optimizing  ML  algorithms  and  the  underlying  system  components  (hardware,  software,  schedulers)  by  considering  their  intricate  interactions.  This  co-design  philosophy  allows  us  to  overcome  the  limitations  of  optimizing  algorithms  and  systems  in  isolation,  unlocking  significant  efficiency  gains.We  demonstrate  the  efficacy  of  this  approach  through  three  core  contributions.  First,  OASIS,  a  system  for  collaborative  neural-enhanced  mobile  video  streaming.  It  addresses  the  challenge  of  running  demanding  super-resolution  models  on  resource-constrained  mobile  devices  with  varying  network  conditions.  OASIS  co-designs  adaptive  bitrate/model  selection  algorithms  with  a  cross-device  collaborative  inference  mechanism,  enabling  multiple  nearby  devices  to  efficiently  share  computation  resources,  significantly  improving  video  Quality  of  Experience  (QoE)  and  reducing  energy  consumption.  Second,  Cake,  a  system  optimizing  Large  Language  Model  (LLM)  inference  latency,  specifically  targeting  the  Key-Value  (KV)  cache  generation  bottleneck  during  long-context  prefill.  It  tackles  the  compute  vs.  I/O  trade-off  in  heterogeneous  memory  systems  by  co-designing  a  bidirectional  parallel  scheduling  strategy  that  dynamically  balances  computing  the  KV  cache  on  GPUs  and  loading  it  from  storage,  substantially  reducing  Time-To-First-Token  (TTFT).  Third,  Plato,  a  semantic-aware  parallel  decoding  framework  for  LLMs  that  addresses  the  inefficiency  of  autoregressive  decoding  while  maintaining  answer  quality.  Plato  co-designs  algorithmic  components  (organizing  sub-problems  into  dependency  graphs  based  on  logical  relationships)  with  system  optimizations  (pipelined  execution  and  KV  cache  reuse).  Our  evaluations  show  that  Plato  improves  throughput  by  up  to  68%  over  autoregressive  decoding  while  maintaining  a  90%  quality  net-win  rate  compared  to  other  parallel  decoding  methods  like  Skeleton-of-Thought.Collectively,  these  contributions  illustrate  that  principled  algorithm-system  co-design  is  a  powerful  and  versatile  methodology  for  building  efficient  ML  inference  systems  that  can  effectively  scale  across  diverse  and  heterogeneous  resources,  optimizing  for  latency  and  throughput  from  the  edge  to  the  datacenter.
■590    ▼aSchool  code:  0127.
■650  4▼aComputer  engineering
■650  4▼aComputer  science
■653    ▼aMachine  learning  systems
■653    ▼aLarge  Language  Models
■653    ▼aEnergy  consumption
■690    ▼a0984
■690    ▼a0464
■690    ▼a0800
■71020▼aUniversity  of  Michigan▼bComputer  Science  &  Engineering.
■7730  ▼tDissertations  Abstracts  International▼g87-03B.
■790    ▼a0127
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17359981▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

Preview

Export

ChatGPT Discussion

AI Recommended Related Books


    New Books MORE
    Statistics for the past 3 years. Go to brief

    Подробнее информация.

    • Бронирование
    • не существует
    • моя папка
    • Первый запрос зрения
    • Non-Book Loan Application
    • Nighttime Book Loan Application
    материал
    Reg No. Количество платежных Местоположение статус Ленд информации
    TF18360 전자도서 대출가능 My Folder 부재도서신고 비도서대출신청 야간 도서대출신청

    * Бронирование доступны в заимствований книги. Чтобы сделать предварительный заказ, пожалуйста, нажмите кнопку бронирование

    Books borrowed together with this book

    Related Popular Books

    Available after logging in.