Numerical Algorithms for Modern Parallel Computer Architectures de Martin Schultz

Numerical Algorithms for Modern Parallel Computer Architectures

Martin Schultz

en Limba Engleză Paperback – 25 feb 2012

Parallel computers have started to completely revolutionize scientific computation. Articles in this volume represent applied mathematics, computer science, and application aspects of parallel scientific computing. Major advances are discussed dealing with multiprocessor architectures, parallel algorithm development and analysis, parallel systems and programming languages. The optimization of the application of massively parallel architectures to real world problems will provide the impetus for the development of entirely new approaches to these technical situations.

Citește tot Restrânge

Cuprins

Parallelization of Physical Systems on the BBN Butterfly: Some Examples.- Solution of Large Scale Engineering Problems Using a Loosely Coupled Array of Processors.- A Look at the Evolution of Mathematical Software for Dense Matrix Problems over the Past Fifteen Years.- A Graphical Approach to Load Balancing and Sparse Matrix Vector Multiplication on the Hypercube.- A Review of Automatic Load Balancing and Decomposition Methods for the Hypercube.- Adaptive Methods for Hyperbolic Problems on Local Memory Parallel Processors.- Systolic Algorithms for the Parallel Solution of Dense Symmetric Positive-Definite Toeplitz Systems.- Ensemble Architectures and their Algorithms: An Overview.- Problem Decomposition and Communication Tradeoffs in a Shared-Memory Multiprocessor.- Domain Decomposition Preconditioners for Elliptic Problems in Two and Three Dimensions.- Methods for Automated Problem Mapping.- Block Algorithms for Parallel Machines.- The LCAP System: An Example of a Multicomputer.- A Block QR Factorization Scheme for Loosely Coupled Systems of Array Processors.