Analytical modelling in parallel and distributed computing /

Examines complex performance evaluation of various typical parallel algorithms (shared memory, distributed memory) and their practical implementations. Includes real application examples we demonstrate the various influences during the process of modelling and performanceevaluation and the consequen...

Descripción completa

Detalles Bibliográficos
Clasificación:Libro Electrónico
Autores principales: Hanuliak, Peter (Autor), Hanuliak, Michal (Autor)
Formato: Electrónico eBook
Idioma:Inglés
Publicado: Oxford [England] : Chartridge Books Oxford, 2014.
Temas:
Parallel processing (Electronic computers)
Electronic data processing > Distributed processing.
Parallélisme (Informatique)
Traitement réparti.
COMPUTERS > Computer Literacy.
COMPUTERS > Computer Science.
COMPUTERS > Data Processing.
COMPUTERS > Hardware > General.
COMPUTERS > Information Technology.
COMPUTERS > Machine Theory.
COMPUTERS > Reference.
Acceso en línea:Texto completo
Tabla de Contenidos:
  • Cover
  • Halftitle
  • Title
  • Copyright
  • Contents
  • Preview
  • Acknowledgements
  • Part I. Parallel Computing
  • Introduction
  • Developing periods in parallel computing
  • 1 Modelling of Parallel Computers and Algorithms
  • Model construction
  • 2 Parallel Computers
  • Classification
  • Architectures of parallel computers
  • Symmetrical multiprocessor system
  • Network of workstations
  • Grid systems
  • Conventional HPC environment versus Grid environments
  • Integration of parallel computers
  • Metacomputing
  • Modelling of parallel computers including communication networks3 Parallel Algorithms
  • Introduction
  • Parallel processes
  • Classification of PAs
  • Parallel algorithms with shared memory
  • Parallel algorithms with distributed memory
  • Developing parallel algorithms
  • Decomposition strategies
  • Natural parallel decomposition
  • Domain decomposition
  • Functional decomposition
  • Mapping
  • Inter process communication
  • Inter process communication in shared memory
  • Inter process communication in distributed memory
  • Performance tuning
  • 4 Parallel Program Developing StandardsParallel programming languages
  • Open MP standard
  • Open MP threads
  • Problem decomposition
  • MPI API standard
  • MPI parallel algorithms
  • Task Groups
  • Communicators
  • The order of tasks
  • Collective MPI commands
  • Synchronisation mechanisms
  • Conditional synchronisation
  • Rendezvous
  • Synchronisation command barriers
  • MPI collective communication mechanisms
  • Data scattering collective communication commands
  • Java
  • 5 Parallel Computing Models
  • SPMD model of parallel computation
  • Fixed (atomic) networkPRAM model
  • Fixed communication model GRAM
  • Flexible models
  • Flexible GRAM model
  • The BSP model
  • Computational model MPMD functionality of all system resources
  • Load of communication network
  • 6 The Role of Performance
  • Performance evaluation methods
  • Analytic techniques
  • Asymptotic (order) analysis
  • Application of queuing theory systems
  • Kendall classification
  • The simulation method
  • Experimental measurement
  • Part II. Theoretical aspects of PA
  • 7 Performance Modelling of Parallel Algorithms
  • Speed upEfficiency
  • Isoefficiency
  • Complex performance evaluation
  • Conclusion and perspectives
  • 8 Modelling in Parallel Algorithms
  • Latencies of PA
  • Part III. Applied Parallel Algorithms
  • 9 Numerical Integration
  • Decomposition model
  • Mapping of parallel processes
  • Performance optimisation
  • Chosen illustration results
  • 10 Synchronous Matrix Multiplication
  • The systolic matrix multiplier
  • Instruction systolic array matrix multiplier
  • ISA matrix multiplier
  • Dataflow matrix multiplication
  • Wave front matrix multiplier

Ejemplares similares