Third International Workshop on
High-Level Parallel Programming Models
and Supportive Environments
(HIPS'98)

IPDPS 1998 Workshop

Preface:

HIPS'98, held in conjunction with the 12th International Parallel Processing Symposium & 9th Symposium on Parallel and Distributed Processing (IPPS/SPDP'98) in Orlando, Florida - USA, was the third in a series of workshops on "High-Level Parallel Programming Models and Supportive Environments". It provided a forum for researchers from industry and academia to discuss recent developments and results on one of the prevailing themes of parallel programming: the design of parallel programming models and languages that (1) are at a high conceptual level, providing more convenience and productivity to the programmer than message-passing models, and that (2) can be implemented efficiently and supported by powerful tools, so that truly high-performance parallel code can be produced by the programming system and the programmer.

Nine high-quality papers were selected for presentation at the one-day workshop and for publication in these proceedings. In addition, an article from John Carter, University of Utah, USA, is included presenting more detailed information from the area covered by his invited presentation on "Distributed Shared Memory: Past, Present, and Future".

While Session 1 gave an overview of the area of distributed shared memory systems, Session 2 covered novel implementations of software DSM systems.

Session 3 was devoted to data parallel programming. It included presentations on language extensions of HPF, performance models for data parallel programming, and on supportive environments for building tools for data parallel languages.

In Session 4 optimization and implementation techniques for high-level programming models were presented. It covered techniques for the efficient implementation of synchronization in the context of shared objects, as well as compiler transformations for high-level programming of shared memory and distributed memory target machines.

In summary, the diverse collection of papers made up for an interesting and stimulating workshop and once again provided evidence that high-level parallel programming is still a widely open and active research topic. Thus, this series of workshops will certainly be continued (in conjunction with IPPS) in the years to come.

I thank all the researchers for their contributions to the workshop and the Program Committee members and reviewers for their time and expertise, which helped to make HIPS‘98 become reality and a successful event. I would like to thank the organizers of the 12th International Parallel Processing Symposium and of the 9th Symposium on Parallel and Distributed Processing for inviting us to organize the workshop and for taking care of the local arrangements.

Michael Gerndt, Workshop Chair
Research Centre Julich
Central Institute for Applied Mathematics
Germany

Program Committee:

Arndt Bode, Technische Universitat Munchen, Germany
Helmar Burkhart, Universitat Basel, Switzerland
John Carter, University of Utah, USA
Michel Cosnard, Ecole Normale Superieure de Lyon, France
Karsten Decker, Swiss Center for Scientific Computing, Switzerland
Dennis Gannon, Indiana University, USA
Michael Gerndt, Research Centre Julich, Germany
Hermann Hellwagner, Technische Universitat Munchen, Germany
Francois Irigoin, Ecole des Mines de Paris, France
Vijay Karamcheti, University of Illinois at Urbana-Champaign, USA
Pete Keleher, University of Maryland, USA
Ulrich Kremer, Rutgers University, USA
Ron Perrott, Queens University Belfast, United Kingdom
Thierry Priol, INRIA, France
Klaus E. Schauser, University of California at Santa Barbara, USA
Domenico Talia, ISI-CNR, Italy
Hans P. Zima, University of Vienna, Austria
Invited Presentation
Making DSM Simple, Yet Efficient
J. Carter

Distributed Shared Memory
Constructive and Adaptable Distributed Shared Memory
Jordi Bataller, Jose M. Bernabeu-Auban

ViC*: A Compiler for Virtual-Memory C*
Alex Colvin, Thomas H. Cormen

Parallel and Distributed Programming with Pthreads and Rthreads
Bernd Dreier, Markus Zahn, Theo Ungerer

Data Parallelism
Language Bindings for a Data-Parallel Runtime
Bryan Carpenter, Geoffrey Fox, Donald Leskiw, Xinying Li, Yuhong Wen, Guansong Zhang

ZPL's WYSIWYG Performance Model
Bradford L. Chamberlain, Sung-Eun Choi, E. Christopher Lewis, Calvin Lin, Lawrence Snyder, W. Derrick Weathersby

A Graph-Based Framework for the Definition of Tools Dealing with Sparse and Irregular Distributed Data Structures
Serge Chaumette, Jean-Michel Lepine, Franck Rubi

Implementation and Optimization Techniques
Implementing Automatic Coordination on Networks of Workstations
Christian Weiss, Jurgen Knopp, Hermann Hellwagner

Improving Performance of Multi-Dimensional Array Redistribution on Distributed Memory Machines
Minyi Guo, Yoshiyuki Yamashita, Ikuo Nakata

Further Results for Improving Loop Interchange in Non-Adjacent and Imperfectly Nested Loops
T.-Ch. Huang, Ch.-M. Yang