EuroPar '97
Uni-Passau FMI Lehrstuhl für Programmierung europar

Workshops 04 + 08 + 13

Workshop 04: Distributed Systems and Algorithms

Programme Committee:

Keith Marzullo, UC San Diego, USA, Global Chair
Friedemann Mattern, TU Darmstadt, Germany, Local Chair
Ozalp Babaoglu, University of Bologna, Italy, Vice-Chair
Fran Berman, UC San Diego, USA, Vice-Chair
Frans Kaashoek, Massachusetts Institute of Technology, USA, Vice-Chair

Description:

Interprocessor communication has become fast enough that distributed systems can be used to solve highly parallel problems in addition to more traditional distributed problems (such as client-server applications). These distributed systems range from a local area network of homogeneous workstations to coordinated heterogenous workstations and supercomputers. Algorithmic and architectural solutions to problems from the fields of distributed and parallel processing (as well as new solutions) can often be applied or adapted to these kinds of systems. Typical examples are implementations of shared memory abstractions on top of message-passing systems, scheduling parallel applications on distributed heterogeneous systems, mechanism and abstractions for fault tolerance, and algorithms to provide elementary system functions and services. This workshop aims at these and other themes mentioned below. Presentations of distributed applications (e.g., databases, cooperative processing, metacomputing, real-time issues) that illustrate pertinent concepts are also welcome. We generally encourage papers that emphasize the junction between distributed computing and parallelism.

Topics of interest include:

Workshop 08: Parallel Discrete Algorithms

Programme Committee:

Cynthia Phillips, Sandia National Laboratories, USA, Global Chair
Michael Kaufmann, University of Tübingen, Germany, Local Chair
Paul Spirakis, Computer Technology Institute, Greece, Vice-Chair
Shang-Hua Teng, University of Minnesota, USA, Vice-Chair

Description:

The goal of this workshop is to provide a forum for advances in the state of the art in parallel discrete algorithms. Parallel algorithms from all discrete areas are appropriate, including discrete optimization (for both specific problems and more generic methods such as parallel search), pattern recognition and image processing, discrete simulation, computational geometry, and graph theory. Special consideration will be given to scalable, practical algorithms. Algorithmic strategies for abstract parallel models such as LogP, BSP, or PRAMs are encouraged, as well as algorithmic tuning for specific architectures, and experimental analysis which provides insight into algorithmic choice. There is a strong relation to workshop 13, but we focus on specific algorithms rather than on models and methods.

Topics of interest include:

Workshop 13: Theory and Models of Parallel Computation

Programme Committee:

Clyde Kruskal, University of Maryland, USA, Global Chair
Klaus-Jörn Lange, University of Tübingen, Germany, Local Chair
Bill McColl, Oxford University, UK, Vice-Chair
Friedhelm Meyer auf der Heide, University of Paderborn, Germany, Vice-Chair

Description:

This workshop intends to discuss the current state of parallel computational and cost models, algorithms and complexity. An emphasis is on research that takes into account properties of real machines like communication cost and asynchrony. Famous examples are BSP and LogP. Developments, comparisons of, and algorithms for such models as well as studies of complexity issues are welcome. There is a strong relation to workshop #08, but we focus on models and methods rather than on specific algorithms.

Topics of interest include:


(C)opyright by University of Passau, Sven Anders 14.05.1997