## Vector and Parallel Processing - VECPAR'98: Third International Conference Porto, Portugal, June 21-23, 1998 Selected Papers and Invited TalksThis book stands as the visible mark of VECPAR'98 – 3rd International Meeting on Vector and Parallel Processing, which was held in Porto (Portugal) from 21 to 23 June 1998. VECPAR'98 was the third of the VECPAR series of conferences initiatedin1993and organisedby FEUP, the FacultyofEngineering of the University of Porto. The conference programme comprised a total of 6 invited talks, 66 c- tributedpapers and18posters. Thecontributed papers andposters were selected from 120 extended abstracts originating from 27 countries. Outline of the book The book, with 7 chapters, contains 41 contributed papers and 6 invited talks. The 41 papers included in these proceedings result from the reviewing of all papers presented at the conference. Each of the ?rst 6 chapters includes 1 of the 6 invited talks of the conference, and related papers. Chapters 1, 2, 4, 5 and 7 are initiated by an introductory text providing the reader with a guide to the chapter contents. Chapter 1 is on numerical algebra.It begins with an introductory text by - cente Hern´andez, followedby the invited talk by Gene Golub, entitled Some - usual Eigenvalue Problems.The remaining11contributed articles in thischapter deal either with large scale eigenvalue problems or with linear system problems. Computational?uiddynamicsandcrashandstructural analysiswere brought under the same chapter and that is Chapter 2, which contains the invited talk byTimothyBarth, entitledParallel Domain Decomposition Pre-conditioning for Computational Fluid Dynamics, plus 8 contributed papers. Timothy Barth also authors the introductory text to the chapter. |

### What people are saying - Write a review

We haven't found any reviews in the usual places.

### Contents

Eigenvalue Problems and Solution of Linear Systems | 1 |

Some Unusual Eigenvalue Problems Invited Talk | 4 |

Multisweep Algorithms for the Symmetric Eigenproblem | 20 |

A Unified Approach to Parallel BlockJacobi Methods for the Symmetric Eigenvalue Problem | 29 |

A Parallel Implementation Using the Implicitly Restarted Arnoldi Method | 43 |

Parallel JacobiDavidson for Solving Generalized Eigenvalue Problems | 58 |

Parallel Preconditioned Solvers for Large Sparse Hermitian Eigenproblems | 71 |

Solving Eigenvalue Problems on Networks of Processors | 85 |

Automatic Detection of Parallel Program Performance Problems | 365 |

Behavioural Analysis Methodology Oriented to Configuration of Parallel RealTime and Embedded Systems | 378 |

Spatial Data Locality with Respect to Degree of Parallelism in ProcessorandMemory Hierarchies | 396 |

Partitioning Regular Domains on Modern Parallel Computers | 411 |

New Access Order to Reduce InterVectorConflicts | 425 |

Registers Size Inﬂuence on Vector Architectures | 439 |

Limits of Instruction Level Parallelism with Data Value Speculation | 452 |

High Performance Cache Management for Parallel File Systems | 466 |

Solving LargeScale Eigenvalue Problems on Vector Parallel Processors | 100 |

Direct Linear Solvers for Vector and Parallel Computers | 114 |

Parallel Preconditioners for Solving Nonsymmetric Linear Systems | 128 |

Synchronous and Asynchronous Parallel Algorithms with Overlap for Almost Linear Systems | 142 |

A ClientServer Model for Interactive Large Scale Scientific Computation | 156 |

Computational Fluid Dynamics Structural Analysis and Mesh Partioning Techniques | 171 |

Parallel DomainDecomposition Preconditioning for Computational Fluid Dynamics Invited Talk | 176 |

Influence of the Discretization Scheme on the Parallel Efficiency of a Code for the Modelling of a Utility Boiler | 203 |

Parallel 3D Airﬂow Simulation on Workstation Clusters | 215 |

Resolving the Inertial Subrange of the Kolmogorov Spectrum | 227 |

The Study of a Parallel Algorithm Using the Laminar BackwardFacing Step Flow as a Test Case | 238 |

A Low Cost Distributed System for FEM Parallel Structural Analysis | 250 |

Dynamic Load Balancing in Crashworthiness Simulation | 263 |

Some Concepts of the Software Package FEAST | 271 |

Aspect Ratio | 285 |

Computing in Education Parallel and Distributed Computing in Education Invited Talk | 301 |

Computer Organisation Programming and Benchmarking | 331 |

Past and Next 10 Years Invited Talk | 334 |

A Systolic Algorithm for the Factorisation of Matrices Arising in the Field of Hydrodynamics | 355 |

Using Synthetic Workloads for Parallel Task Scheduling Improvement Analysis | 480 |

Dynamic Routing Balancing in Parallel Computer Interconnection Networks | 494 |

AlgorithmDependant Method to Determine the Optimal Number of Computers in Parallel Virtual Machines | 508 |

A Way to be Efficient | 522 |

A Performance Analysis of the SGI Origin2000 | 534 |

An ISA Comparison Between Superscalar and Vector Processors | 548 |

Image Analysis and Synthesis | 561 |

Invited Talk | 563 |

Parallel Implementations of Morphological Connected Operators Based on Irregular Data Structures | 579 |

Parallel Database Servers | 593 |

Nonlinear Problems | 622 |

A Parallel NBody Integrator Using MPI | 627 |

A Parallelisation Strategy for Power Systems Composite Reliability Evaluation Best Student Paper Award | 640 |

First Prize | 652 |

Calculations | 666 |

Simulating Magnetised Plasma with the Versatile Advection Code | 680 |

Parallel Genetic Algorithms for Hypercube Machines | 691 |

705 | |

### Other editions - View all

### Common terms and phrases

analysis application bandwidth behaviour benchmark block buﬀer cache cache coherence calculation client communication conﬁguration conﬂicts convergence corresponding cost Cray T3E data value speculation decomposition deﬁned diﬀerent distributed memory domain Dongarra dynamic eﬀect eﬃciency eigenvalue problem eigenvectors elements equations execution factorisation ﬁeld ﬁgure ﬁle ﬁrst ﬂow function global graph grid high performance implementation inﬂuence instructions interface LAPACK latency linear systems load balancing machine macro-activity matrix matrix-vector memory modules mesh message passing Metapath method MUSCL scheme nodes number of processors object obtained ODBMS operations optimisation overlapping parallel algorithm parallel computers parallel program parallelisation parameters partitioning pixels preconditioner protocol radiosity reduce scalability ScaLAPACK scheme Schur complement sequential signiﬁcant simulation solution solve solver sparse sparse matrix speciﬁc speedup step structure subdomains Supernode superscalar symmetric Table tabu search technique tion traﬃc updated vector processors vector register vector streams vectorisation