본문

서브메뉴

Fast, Accurate and Scalable Time Evolution in Quantum and Continuum Systems With Exponential Propagators
Fast, Accurate and Scalable Time Evolution in Quantum and Continuum Systems With Exponenti...
Fast, Accurate and Scalable Time Evolution in Quantum and Continuum Systems With Exponential Propagators

Detailed Information

자료유형  
 학위논문 서양
최종처리일시  
20260202105225
ISBN  
9798291566664
DDC  
519
저자명  
Pari, Paavai.
서명/저자  
Fast, Accurate and Scalable Time Evolution in Quantum and Continuum Systems With Exponential Propagators
발행사항  
[Sl] : University of Michigan, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
208 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-03, Section: B.
주기사항  
Advisor: Gavini, Vikram.
학위논문주기  
Thesis (Ph.D.)--University of Michigan, 2025.
초록/해제  
요약We develop computationally efficient, accurate, stable and scalable time propagators for applications in elastodynamics and quantum dynamics using exponential propagators. In elastodynamics we recast the second-order dynamical equation of elastodynamics into an equivalent first-order system of coupled equations, so as to express the solution in the form of a Magnus expansion. With any spatial discretization, it entails computing the exponential of a matrix acting upon a vector. We employ an adaptive Krylov subspace approach to inexpensively and accurately evaluate the action of the exponential matrix on a vector. In particular, we use an apriori error estimate to predict the optimal Krylov subspace size required for each time-step size. We show that the Magnus expansion truncated after its first term provides quadratic and superquadratic convergence in the time-step size for nonlinear and linear elastodynamics, respectively. We demonstrate the accuracy and efficiency of the proposed method for linear and nonlinear benchmark systems. For a desired accuracy in energy, displacement, and velocity, our method allows for 10 − 100x larger time-step sizes than conventional time-marching schemes such as Newmark-β method. Computationally, it translates to a ∼1000x and ∼10 − 100x speed-up over conventional time-marching schemes for linear and nonlinear elastodynamics, respectively. To advance the frontiers of quantum dynamics simulations, we have developed TDDFT-FE (a real-time time-dependent density functional theory (RT-TDDFT) capability) within the DFT-FE framework. This work presents the theory, implementation, and validation of TDDFT-FE, establishing a scalable and robust platform for accurate and stable quantum dynamics simulations on large-scale systems. By harnessing the inherent first-order nature of the time-dependent Schr¨odinger equation, TDDFT-FE employs exponential propagators to enable efficient, unitary, and precise time evolution. In this work, we develop and systematically benchmark advanced numerical strategies including Chebyshev polynomial expansion, Lanczos, and Two-Pass Lanczos methods for efficient evaluation of the matrix exponential. We implement a hybrid CPU-GPU parallelization strategy coupled with band-parallelism, achieving maximal utilization of high-performance computing resources and seamless scalability to systems with thousands of electrons. Our numerical studies demonstrate the superior accuracy of TDDFT-FE compared to a state-of-the-art real-space TDDFT code, with significant performance gains in both computational efficiency and degrees of freedom handled. We showcase TDDFT-FE's broad capability across diverse systems, from organic molecules to large nanoclusters with tens of thousands of electrons, including simulations in the regime of strong nonlinear response. Detailed convergence studies, CPU-GPU performance analyses, and extreme-scaling demonstrations establish TDDFT-FE as a robust and scalable platform for large-scale, long-time quantum dynamics simulations. By enabling extremely fast, accurate, and stable TDDFT calculations, TDDFT-FE not only provides direct access to ultrafast phenomena on attosecond timescales, but also creates a critical foundation for next-generation materials discovery, chemical innovation, and the development of time-dependent exchange-correlation functionals essential for advancing predictive quantum dynamics.
일반주제명  
Applied mathematics
일반주제명  
Materials science
일반주제명  
Mechanical engineering
일반주제명  
Engineering
일반주제명  
Quantum physics
키워드  
Exponential integrators
키워드  
Elastodynamics
키워드  
Time dependent density functional theory
키워드  
Quantum dynamics
키워드  
Magnus propagators
키워드  
Finite element method
기타저자  
University of Michigan Mechanical Engineering
기본자료저록  
Dissertations Abstracts International. 87-03B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017359851
■00520260202105225
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798291566664
■035    ▼a(MiAaPQ)AAI32271843
■035    ▼a(MiAaPQ)umichrackham006433
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a519
■1001  ▼aPari,  Paavai.
■24510▼aFast,  Accurate  and  Scalable  Time  Evolution  in  Quantum  and  Continuum  Systems  With  Exponential  Propagators
■260    ▼a[Sl]▼bUniversity  of  Michigan▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a208  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-03,  Section:  B.
■500    ▼aAdvisor:  Gavini,  Vikram.
■5021  ▼aThesis  (Ph.D.)--University  of  Michigan,  2025.
■520    ▼aWe  develop  computationally  efficient,  accurate,  stable  and  scalable  time  propagators  for  applications  in  elastodynamics  and  quantum  dynamics  using  exponential  propagators.  In  elastodynamics  we  recast  the  second-order  dynamical  equation  of  elastodynamics  into  an  equivalent  first-order  system  of  coupled  equations,  so  as  to  express  the  solution  in  the  form  of  a  Magnus  expansion.  With  any  spatial  discretization,  it  entails  computing  the  exponential  of  a  matrix  acting  upon  a  vector.  We  employ  an  adaptive  Krylov  subspace  approach  to  inexpensively  and  accurately  evaluate  the  action  of  the  exponential  matrix  on  a  vector.  In  particular,  we  use  an  apriori  error  estimate  to  predict  the  optimal  Krylov  subspace  size  required  for  each  time-step  size.  We  show  that  the  Magnus  expansion  truncated  after  its  first  term  provides  quadratic  and  superquadratic  convergence  in  the  time-step  size  for  nonlinear  and  linear  elastodynamics,  respectively.  We  demonstrate  the  accuracy  and  efficiency  of  the  proposed  method  for  linear  and  nonlinear  benchmark  systems.  For  a  desired  accuracy  in  energy,  displacement,  and  velocity,  our  method  allows  for  10  −  100x  larger  time-step  sizes  than  conventional  time-marching  schemes  such  as  Newmark-β  method.  Computationally,  it  translates  to  a  ∼1000x  and  ∼10  −  100x  speed-up  over  conventional  time-marching  schemes  for  linear  and  nonlinear  elastodynamics,  respectively. To  advance  the  frontiers  of  quantum  dynamics  simulations,  we  have  developed  TDDFT-FE  (a  real-time  time-dependent  density  functional  theory  (RT-TDDFT)  capability)  within  the  DFT-FE  framework.  This  work  presents  the  theory,  implementation,  and  validation  of  TDDFT-FE,  establishing  a  scalable  and  robust  platform  for  accurate  and  stable  quantum  dynamics  simulations  on  large-scale  systems.  By  harnessing  the  inherent  first-order  nature  of  the  time-dependent  Schr¨odinger  equation,  TDDFT-FE  employs  exponential  propagators  to  enable  efficient,  unitary,  and  precise  time  evolution.  In  this  work,  we  develop  and  systematically  benchmark  advanced  numerical  strategies  including  Chebyshev  polynomial  expansion,  Lanczos,  and  Two-Pass  Lanczos  methods  for  efficient  evaluation  of  the  matrix  exponential.  We  implement  a  hybrid  CPU-GPU  parallelization  strategy  coupled  with  band-parallelism,  achieving  maximal  utilization  of  high-performance  computing  resources  and  seamless  scalability  to  systems  with  thousands  of  electrons.  Our  numerical  studies  demonstrate  the  superior  accuracy  of  TDDFT-FE  compared  to  a  state-of-the-art  real-space  TDDFT  code,  with  significant  performance  gains  in  both  computational  efficiency  and  degrees  of  freedom  handled.  We  showcase  TDDFT-FE's  broad  capability  across  diverse  systems,  from  organic  molecules  to  large  nanoclusters  with  tens  of  thousands  of  electrons,  including  simulations  in  the  regime  of  strong  nonlinear  response.  Detailed  convergence  studies,  CPU-GPU  performance  analyses,  and  extreme-scaling  demonstrations  establish  TDDFT-FE  as  a  robust  and  scalable  platform  for  large-scale,  long-time  quantum  dynamics  simulations.  By  enabling  extremely  fast,  accurate,  and  stable  TDDFT  calculations,  TDDFT-FE  not  only  provides  direct  access  to  ultrafast  phenomena  on  attosecond  timescales,  but  also  creates  a  critical  foundation  for  next-generation  materials  discovery,  chemical  innovation,  and  the  development  of  time-dependent  exchange-correlation  functionals  essential  for  advancing  predictive  quantum  dynamics.
■590    ▼aSchool  code:  0127.
■650  4▼aApplied  mathematics
■650  4▼aMaterials  science
■650  4▼aMechanical  engineering
■650  4▼aEngineering
■650  4▼aQuantum  physics
■653    ▼aExponential  integrators
■653    ▼aElastodynamics
■653    ▼aTime  dependent  density  functional  theory
■653    ▼aQuantum  dynamics
■653    ▼aMagnus  propagators
■653    ▼aFinite  element  method
■690    ▼a0364
■690    ▼a0548
■690    ▼a0794
■690    ▼a0599
■690    ▼a0537
■71020▼aUniversity  of  Michigan▼bMechanical  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=T17359851▼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

    高级搜索信息

    • 预订
    • 不存在
    • 我的文件夹
    • 第一种观点
    • Non-Book Loan Application
    • Nighttime Book Loan Application
    材料
    注册编号 呼叫号码. 收藏 状态 借信息.
    TF15489 전자도서 대출가능 My Folder 부재도서신고 비도서대출신청 야간 도서대출신청

    *保留在借用的书可用。预订,请点击预订按钮

    Books borrowed together with this book

    Related Popular Books

    Available after logging in.