Vector and Parallel Processing - VECPAR 2000

This book is the ?nal outcome of VECPAR 2000 - 4th International Meeting on Vector and Parallel Processing. VECPAR constitutes a series of conferences, which have been organized by the Faculty of Engineering of the University of Porto since 1993, with the main objective of disseminating new knowledge on parallel computing. Readership of This Book The book is aimed at an audience of researchers and graduate students in a broad range of scienti?c areas, including not only computer science, but also applied mathematics and numerical analysis, physics, and engineering. Book Plan From a total of 66 papers selected on the basis of extended abstracts for p- sentation at the conference, a subset of 34 papers were chosen during a second review process leading to their inclusion in the book, together with the invited talks. The book contains a total of 40 papers organized into 6 chapters, where each may appeal to people in di?erent but still related scienti?c areas. All ch- ters, with the exception of Chapter 6, are initiated by a short text, providing a quick overview of the organization and papers in the chapter. The 13 papers in Chapter 1 cover the aspects related to the use of multiple processors. Operating systems, languages and software tools for scheduling, and code transformation are the topics included in this chapter, initiated by the talk on computing over the Internet, entitled Grid Computing,byIan Foster.

Inhalt

Computational Grids, Languages, and Tools in Multiplatform Environments.- Computational Grids.- The Distributed Engineering Framework TENT.- A Parallel VRML97 Server Based on Active Objects.- Dynamic Load Balancing Model: Preliminary Results for Parallel Pseudo-search Engine Indexers/Crawler Mechanisms Using MPI and Genetic Programming.- Implementing and Analysing an Effective Explicit Coscheduling Algorithm on a NOW.- Enhancing Parallel Multimedia Servers through New Hierarchical Disk Scheduling Algorithms.- Suboptimal Communication Schedule for GEN_BLOCK Redistribution.- A SCOOPP Evaluation on Packing Parallel Objects in Run-Time.- Measuring the Performance Impact of SP-Restricted Programming in Shared-Memory Machines.- A Neural Network Based Tool for Semi-automatic Code Transformation.- A Platform Independent Parallelising Tool Based on Graph Theoretic Models.- Improving the Performance of Heterogeneous DSMs via Multithreading.- Value Prediction as a Cost-Effective Solution to Improve Embedded Processors Performance.- Cellular Automata and Applications in Computational Physics.- Cellular Automata: Applications.- The Role of Parallel Cellular Programming in Computational Science.- Optimization with Parallel Computing.- Parallelization of a Density Functional Program for Monte-Carlo Simulation of Large Molecules.- Power System Reliability by Sequential Monte Carlo Simulation on Multicomputer Platforms.- A Novel Algorithm for the Numerical Simulation of Collision-Free Plasma-Vlasov Hybrid Simulation.- An Efficient Parallel Algorithm for the Numerical Solution of Schrödinger Equation.- Linear and Non-linear Algebra.- Parallel Branch-and-Bound for Chemical Engineering Applications: Load Balancing and Scheduling Issues.- A Parallel Implementation of an Interior-Point Algorithm for Multicommodity Network Flows.- A Parallel Algorithm for Solving the Toeplitz Least Squares Problem.- An Index Domain for Adaptive Multi-grid Methods.- Parallelization of a Recursive Decoupling Method for Solving Tridiagonal Linear Systems on Distributed Memory Computer.- A New Parallel Approach to the Toeplitz Inverse Eigenproblem Using Newton-like Methods.- An Efficient Parallel Algorithm for the Symmetric Tridiagonal Eigenvalue Problem.- Non-stationary Parallel Newton Iterative Methods for Nonlinear Problems.- Parallel Pole Assignment of Single-Input Systems.- Solving the Generalized Sylvester Equation with a Systolic Library.- Imaging.- Thirty Years of Parallel Image Processing.- Parallel Image Processing System on a Cluster of Personal Computers.- Synchronous Non-local Image Processing on Orthogonal Multiprocessor Systems.- Parallel Implementation of a Track Recognition System Using Hough Transform.- Finite/Discrete Elements in Engineering Applications.- Finite/Discrete Element Analysis of Multi-fracture and Multi-contact Phenomena.- Parallel Edge-Based Finite Element Techniques for Nonlinear Solid Mechanics.- High Level Parallelization of a 3D Electromagnetic Simulation Code with Irregular Communication Patterns.- Parallel Algorithm for Fast Cloth Simulation.- A Parallel Algorithm for the Simulation of the Dynamic Behaviour of Liquid-Liquid Agitated Columns.- Simulation of Turbulent Flows.- Large-Eddy Simulations of Turbulent Flows, from Desktop to Supercomputer.