Joubert G.R., Nagel W.E., Peters E.J., Walter W.V. Parallel Computing. Software Technology, Algorithms, Architectures and Applications

Elsevier, 2004. — 975 p.Dresden, a city of science and technology, of fine arts and baroque architecture, of education and invention, location of important research institutes and high tech firms in IT and biotechnology, and gateway between Western and Eastern Europe, attracted 175 scientists for the international conference on parallel computing ParCo2003 from 2 to 5 September 2003. It was the tenth in the biannual ParCo series, the longest running European conference series covering all aspects of parallel and high performance computing. ParCo2003 was once again a milestone in gauging the status quo of research and the state of the art in the development and application of parallel and high performance computing techniques, highlighting both current and future trends.
The conference was hosted by the Center of High Performance Computing (ZHR) of the Technical University of Dresden. Since its foundation in 1828, the TU Dresden has undergone tremendous transformations from engineering school to technical school of higher education to full university. Today, the TU Dresden offers a broad range of subjects and specialisations in a wide variety of fields to about 33000 students. In the tradition of many inventions in the development of mechanical calculators and early computers, the Center for High Performance Computing was founded in 1997 and has played an important role in the development of modem methods and tools to support high performance computing at the university and beyond.
Nowadays, many aspects of parallel computing have become part of mainstream computing. It is now commonplace to buy commodity off-the-shelf computers for home and office use that incorporate parallel techniques such as superscalarity, hyper threading, VLIW (Very Long Instruction Word) and even cluster technologies that were considered advanced a mere decade ago. Quite apart from the speed with which new parallel technologies find their way into new products, these developments underline the importance of parallel computing research and development for the advancement of computer science and IT in general. In view of the rapid technology transfer taking place, one could be led to conclude that parallel computing research and development has passed its zenith since it has become standard computing practice.
ParCo2003 showed that such a conclusion is invalid and that many complex research issues remain to be investigated. Thus it is clear and this has been the case for a number of years – that future research in parallel computing will have to concentrate increasingly on all aspects of software engineering. In addition the development of new architectures, especially those based on new technologies such as nanotechnologies, biocomputing, improved methods for performance evaluation, advanced algorithms, etc. must continue to receive appropriate attention.Invited Papers
Parallel Machines and the "Digital Brain" - An Intricate Extrapolation on Occasion of JvN's 100-th Birthday
So Much Data, So Little Time...
Software Technology
On Compiler Support for Mixed Task and Data Parallelism
Distributed Process Networks - Using Half FIFO Queues in COREA
An efficient data race detector backend for DIOTA
Pipelined parallelism for multi-join queries on shared nothing machines
Towards the Hierarchical Group Consistency for DSM systems: an efficient way to share data objects
An operational semantics for skeletons
A Programming Model for Tree Structured Parallel and Distributed Algorithms and its Implementation in a Java Environment
A Rewriting Semantics for an Event-Oriented Functional Parallel Language
RMI-like communication for migratable software components in HARNESS
Semantics of a Functional BSP Language with Imperative Features
The Use of Parallel Genetic Algorithms for Optimization in the Early Design Phases
An Integrated Annotation and Compilation Framework for Task and Data Parallel Programming in Java
On The Use of Java Arrays for Sparse Matrix Computations
A Calculus of Functional BSP Programs with Explicit Substitution
JToe: a Java API for Object Exchange
A Modular Debugging Infrastructure for Parallel Programs
Toward a Distributed Computational Steering Environment based on CORBA
Parallel Decimation of 3D Meshes for Efficient Web-Based Isosurface Extraction
Parallel Programming
MPI on a Virtual Shared Memory
OpenMP vs. MPI on a Shared Memory Multiprocessor
MPI and OpenMP implementations of Branch-and-Bound Skeletons
Parallel Overlapped Block-Matching Motion Compensation Using MPI and OpenMP
A comparison of OpenMP and MPI for neural network simulations on a SunFire 6800
Comparison of Parallel Implementations of Runge-Kutta Solvers: Message Passing vs. Threads
Scheduling
Extending the Divisible Task Model for Workload Balancing in Clusters
The generalized diffusion method for the load balancing problem
Delivering High Performance to Parallel Applications Using Advanced Scheduling
Algorithms
Multilevel Extended Algorithms in Structural Dynamics on Parallel Computers
Parallel Model Reduction of Large-Scale Unstable Systems
Parallel Decomposition Approaches for Training Support Vector Machines
Fast parallel solvers for fourth-order boundary value problems
Parallel Solution of Sparse Eigenproblems by Simultaneous Rayleigh Quotient Optimization with FSAI preconditioning
An Accurate and Efficient Selfverifying Solver for Systems with Banded Coefficient Matrix
3D parallel calculations of dendritic growth with the lattice Boltzmann method
Distributed Negative Cycle Detection Algorithms
A Framework for Seamlessly Making Object Oriented Applications Distributed
Performance Evaluation of Parallel Genetic Algorithms for Optimization Problems of Different Complexity
Extensible and Customizable Just-In-Time Security (JITS) Management of Client-Server Communication in Java
Applications & Simulation
An Object-Oriented Parallel Multidisciplinary Simulation System The SimServer
Computer Simulation of Action Potential Propagation on Cardiae Tissues: An Efficient and Scalable Parallel Approach
MoDySim A parallel dynamical UMTS simulator
apeNEXT: a Multi-TFlops Computer for Elementary Particle Physics
The Parallel Model System LM-MUSCAT for Chemistry-Transport Simulations: Coupling Scheme, Parallelization and Applications
Real-time Visualization of Smoke through Parallelizations
Parallel Simulation of Cavitated Flows in High Pressure Systems
Improvements in black hole detection using parallelism
High Throughput Computing for Neural Network Simulation
Parallel algorithms and data assimilation for hydraulic models
Multimedia Applications
Parallelization of VQ Codebook Generation using Lazy PNN Algorithm
A Scalable Parallel Video Server Based on Autonomous Network-attached Storage
Efficient Parallel Search in Video Databases with Dynamic Feature Extraction
Architectures
Introspection in a Massively Parallel PIM-Based Architecture
Time-Transparent Inter-Processor Connection Reconfiguration in Parallel Systems Based on Multiple Crossbar Switches
SIMD design to solve partial differential equations
Caches
Trade-offs for Skewed-Associative Caches
Cache Memory Behavior of Advanced PDE Solvers
Performance
A Comparative Study of MPI Implementations on a Cluster of SMP Workstations
MARMOT: An MPI Analysis and Checking Tool
BenchIT Performance Measurement and Comparison for Scientific Applications
Performance Issues in the Implementation of the M-VIA Communication Software
Performance and performance counters on the Itanium 2 - A benchmarking case study
On the parallel prediction of the RNA secondary structure
Clusters
MDICE a MatLAB Toolbox for Efficient Cluster Computing
Parallelization of Krylov Subspace Methods in Multiprocessor PC Clusters
First Impressions of Different Parallel Cluster File Systems
Fast Parallel I/O on ParaStation Clusters
PRFX: a runtime library for high performance programming on clusters of SMP nodes
Grids
Experiences about Job Migration on a Dynamic Grid Environment
Security in a Peer-to-Peer Distributed Virtual Environment
A Grid Environment for Diesel Engine Chamber Optimization
A Broker Architecture for Object-Oriented Master/Slave Computing in a Hierarchical Grid System
A framework for experimenting with structured parallel programming environment design
Minisymposium - Grid Computing
Considerations for Resource Brokerage and Scheduling in Grids
Job Description Language and User Interface in a Grid context: The EU DataGrid experience
On Pattern Oriented Software Architecture for the Grid
Minisymposium - Bioinformatics
Green Destiny + mpiBLAST = Bioinfomagic
Parallel Processing on Large Redundant Biological Data Sets: Protein Structures Classification with CEPAR
MDGRAPE-3: A Petafiops Special-Purpose Computer System for Molecular Dynamics Simulations
Structural Protein Interactions: From Months to Minutes
Spatially Realistic Computational Physiology: Past, Present and Future
Cellular automaton modeling of pattern formation in interacting cell systems
Numerical Simulation for eHealth: Grid-enabled Medical Simulation Services
Parallel computing in biomedical research and the search for peta-scale biomedical applications
Minisymposium - Performance Analysis
Big Systems and Big Reliability Challenges
Scalable Performance Analysis ofParalle1 Systems: Concepts and Experiences
CrossWalk: A Tool for Performance Profiling Across the User-Kernel Boundary
Hardware-Counter Based Automatic Performance Analysis of Parallel Programs
Online Performance Observation of Large-Scale Parallel Applications
Deriving analytical models from a limited number of runs
Performance Modeling of HPC Applications
Minisymposium - OpenMP
Thread based OpenMP for nested parallelization
OpenMP on Distributed Memory via Global Arrays
Performance Simulation of a Hybrid OpenMPIMPI Application with HeSSE
An environment for OpenMP code parallelization
Hindrances in OpenMP programming
Wavelet-Based Still Image Coding Standards on SMPs using OpenMP
Minisymposium - Parallel Applications
Parallel Solution of the Bidomain Equations with High Resolutions
Balancing Domain Decomposition Applied to Structural Analysis Problems
Multiperiod Portfolio Management Using Parallel Interior Point Method
Performance of a parallel split operator method for the time dependent Schrodinger equation
Minisymposium - Cluster Computing
Design and implementation of a 512 CPU cluster for general purpose supercomputing
Experiences Parallelizing, Configuring, Monitoring, and Visualizing Applications for Clusters and Multi-Clusters 879
Cluster Computing as a Teaching Tool
Minisymposium - Mobile Agents
Mobile Agents Principles of Operation
Mobile Agent Application Fields
Mobile Agents and Grid Computing
Mobile Agents, Globus and Resource Discovery
A Mobile Agent Tool for Resource Discovery
Mobile Agents and Knowledge Discovery in Ubiquitous Computing