Lecture Notes in Computer Science Edited by G. Goos, J. Hartmanis and J. van Leeuwen

1541

Springer Berlin Heidelberg New York Barcelona Hong Kong London Milan Paris Singapore Tokyo

Bo Kglgstr6m Jack Dongarra Erik Elmroth Jerzy Wagniewski (Eds.)

Applied Parallel Computing Large Scale Scientific and Industrial Problems

4th International Workshop, PARA' 98 Ume~, Sweden, June 14-17, 1998 Proceedings

Springer

Series Editors

Gerhard Goos, Karlsruhe University, Germany Juris Hartmanis, Cornell University, NY, USA Jan van Leeuwen, Utrecht University, The Netherlands

Volume Editors

Bo K~gstrOm Erik Elmroth Ume~t University, Dept. of Computing Science and HPC2N S-901 87 Sweden E-mail: {bo.kagstrom,elmroth} @cs.umu.se

Jack Dongarra University of Tennessee, 107 Ayres Hall Knoxville, TN 37996-1301, USA E-mail: dongarra@cs.utk.edu

Jerzy Wagniewski Danish Computing Centre for Research and Education DTU, UNI C, Bldg. 304, DK-2800 Lyngby, Denmark Jerzy.Wasniewski @ uni-c.dk

Cataloging-in-Publication data applied for

Die Deutsche Bibliothek - CIP-Einheitsaufilahme

Applied parallel computing : large scale scientific and industrial problem ; 4th international workshop ; proceedings / PARA "98, Ume~, Sweden, June 14 - 17, 1998. Bo K~gstr6m ... (ed.). - Berlin ; Heidelberg ; New York , Barcelona ; Hong Kong ; Lo n d o n , Milan ; Paris ; Singapore ; Tokyo : Springer, 1998

(Lecture notes in computer science ; Vol. 1541) ISBN 3-540-65414-3

CR Subject Classification (1998): G.1-2, G.4, F.1-2, D.1-3, J.1

ISSN 0302-9743 ISBN 3-540-65414-3 Springer-Verlag Berlin Heidelberg New York

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the German Copyright Law.

�9 Springer-Verlag Berlin Heidelberg 1998 Printed in Germany

Typesetting: Camera-ready by author SPIN 10693025 06/3142 - 5 4 3 2 1 0 Printed on acid-free paper

Preface

The Fourth International Workshop on Applied Parallel Computing (PARA'98) was held in Ume~, Sweden, June 14-17, 1998. The workshop was organized by the High Performance Computing Center North (HPC2N) and the Department of Computing Science at Ume~ University. The general theme for PARA'98 was Large Scale Scientific and Industrial Problems, focusing on:

- High-performance computing applications in academia and industry, - Tools, languages and environments for high-performance computing, - Scientific visualization and virtual reality applications in academia

and industry, - Future directions in high-performance computing and networking.

The workshop attracted over 140 people representing 18 different countries. PARA'98 was an international forum for idea and competence exchange for specialists in high performance and parallel computing, visualization, and scien- tists from industry and academia involved in solving large scale computational problems. The workshop program included 20 invited presentations and 64 con- tributed presentations that were selected by the PARA'98 steering committee. These proceedings reflect the results of this meeting.

The PARA'98 meeting began with a one-day tutorial followed by a three-day workshop. The tutorial program included three topics: Tools and Languages for High Performance Computing (Dennis Gannon, Indiana University), Projection Based Virtual Environments for Collaboration in Scientific Visualization, Indus- trial Design and Art (Dan Sandin, University of Illinois, Chicago), and Scientific Visualization and Computational Steering, (Chris Johnson, University of Utah). Over 80 people attended the tutorials.

The first three PARA workshops were held at the Technical University of Denmark (DTU), Lyngby (1994, 1995, and 1996). Following PARA'96 an in- ternational steering committee for the PARA meetings was appointed and the committee decided that a workshop should take place every second year in one of the Nordic countries. One important aim of these workshops is to strengthen the ties between HPC centers, academia, and industry in the Nordic countries as well as worldwide. Sweden and Ume~ University organized the 1998 work- shop and the next workshop in the year 2000 will take place at the University of Bergen in Norway.

September 1998 Bo K~gstrSm Jack Dongarra

Erik Elmroth Jerzy Wa~niewski

VI

Organization

PARA'98 was organized by the High Performance Computing Center North (HPC2N) and the Department of Computing Science at Umes University.

Organizing Committee

Conference Chair: Conference Coordinator: Local Organization:

Bo Ks (Umes University, Sweden) Erik Elmroth (Umes University, Sweden) Krister Dackland (Umes University, Sweden) Lena Hellman (Umes University, Sweden) Per Ling (Umes University, Sweden) Mats NylSn (Umes University, Sweden) Peter Poromaa (Umes University, Sweden)

Steering Committee

Petter Bj0rstad Jack Dongarra

BjSrn Engquist

Kristjan Jonasson Bo Ks

Risto Nieminen

Karstein S0rli

Olle Teleman

Jerzy Wa~niewski

University of Bergen (Norway) University of Tennessee and Oak Ridge National Laboratory (USA) PDC, Royal Institute of Technology, Stockholm (Sweden) University of Iceland, Reykjavik (Iceland) Umes University and HPC2N (Sweden), PARA '98 Chairman Helsinki University of Technology, Espoo (Finland) SINTEF, Dept of Industrial Mathematics, Trondheim (Norway) Center for Scientific Computing (CSC), Espoo (Finland) Danish Computing Centre for Research and Education (UNI.C), Lyngby (Denmark), PARA '9~-96 Chairman

Sponsoring Institutions

Swedish Council for High Performance Computing (HPDR) Swedish Natural Science Research Council (NFR) Swedish Research Council for Engineering Sciences (TFR) Umes University (Rector, HPC2N, Department of Computing Science) IBM Sweden

A c k n o w l e d g m e n t s

We acknowledge the enthusiastic work of the steering committee and the local or- ganizing committee. We also acknowledge the following people for the assistance and support in the organization of PARA'98: Inga Bohman, Lena Carneland, Anne-Lie Persson, Inger Sandgren; Anders Backman, Erik Bs TorbjSrn Jo- hansson, Mikael R~nnar,/~ke Sandgren, and BjSrn Torkelsson. PARA'98 would not have been possible without the personal involvement of all these people. We thank Krister Dackland, Isak Jonsson, and Per Ling for their professional assistance in editing the proceedings. Finally, we also would like to thank the sponsoring institutions for their generous financial support.

Table of Contents

Speaker information. 1

Communications Latency Hiding Techniques for a Reconfigurable Optical Interconnect: Benchmark Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

A. A#ahi, N.J. Dimopoulos

Multifrontal Solvers Within the PARASOL Environment . . . . . . . . . . . . . . . . 7 P. Amestoy, I. Duff, J.-Y. L'Excellent

Parallelization of a 3D FD-TD Code for the Maxwell Equations Using MPI 12 U. Andersson

Advanced Calculations and Visualization of Enzymatic Reactions with the Combined Quantum Classical Molecular Dynamics Code . . . . . . . . . . . . . . . 20

P. Bata, P. Grochowski, K. Nowidski, T. Clark, B. Lesyng, J.A. McCammon

Memory Access Profiling Tools for Alpha-based Architectures . . . . . . . . . . . 28 S. M. Balle, S.C. Steely, Jr.

Parallelized Block-Structured Newton-Type Methods in Dynamic Process Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

J. Borchardt

Tuning the Performance of Parallel Programs on NOWs Using Performance Analysis Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

M. Bubak, W. Funika, J. Mogci~ski

Numerical Simulation of 3D Fully Nonlinear Water Waves on Parallel Computers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

X. Cai

Fluctuations in the Defect Creation by Ion Beam Irradiation . . . . . . . . . . . . 56 R. Chakarova, L Pdzsit

Parallelisation of an Industrial Hydrodynamics Application Using the P INEAPL Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

T. Christensen, A.R. Krornmer, J. Larsen, L. SOrensen

Hyper-Rectangle Selection Strategy for Parallel Adaptive Numerical Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

R. Ciegis, R. Sablinskas, J. Wa{niewski

1 Bold style indicates the invited speaker. Underline indicates the speaker.

•

Parallelising ~-hzzy Queries for Spatial Da ta Modelling on a Cray T3D . . . . 76 A. Clematis, A. Coda, M. Spagnuolo, S. SpineUo, T. Sloan

Hyper-Systolic Implementat ion of BLAS-3 Routines on the APE100/Quadr ics Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

M. Coletta, T. Lippert, P. Palazzari

Resource Management for Ultra-scale Computat ional Grid Appl ica t ions . . . 88 K. Czajkowski, I. Foster, C. K e s s e l m a n

A ScaLAPACK-Style Algorithm for Reducing a Regular Matr ix Pair to Block Hessenberg-Triangular Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

K. Dackland, B. Kdgstr5m

Parallel Tight-Binding Molecular Dynamics Code Based on Integrat ion of H P F and Optimized Parallel Libraries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

B. Di Martino, M. Celino, M. Briscolini, L. Colombo, S. Filippone, V. Rosato

Parallel Computat ion of Multidimensional Scattering Wavefunctions for Helmholtz/Schroedinger Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

/t. Edlund, I. Bar-On, U. Peskin

New Serial and Parallel Recursive QR Factorization Algorithms for SMP Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

E. Elmroth, F. Gustavson

Visualization of CFD Computat ions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 J. EngstrSm

Improving the Performance of Scientific Parallel Applications in a Cluster of Workstations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

A. Flores, J.M. Garcla

On the Parallelisation of Non-linear Optimisation Algorithms for Ophthalmical Lens Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

E. Fontdecaba Baig, J.M. Cela Espln, J.C. Diirsteler Lopez

Modelica - - A Language for Equation-Based Physical Modeling and High Performance Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

P. F r i t z s o n

Distributed Georeferring of Remotely Sensed Landsa t -TM Imagery Using MPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

J.D. Garcla-Consuegra, J.A. Gallud, G. Sebastidn

Parallel Test Pat tern Generation Using Circuit Parti t ioning in a Shared-Memory Multiprocessor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

C. Gil, Y. Ortega, J.L. Bernier, M.D. Gil

•

Parallel Adaptive Mesh Refinement for Large Eddy Simulation Using the Finite Element Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

D. Golden, N. Hurley, S. McGrath

WSSMP: A High-Performance Serial and Parallel Symmetric Sparse Linear Solver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

A. G u p t a , M. Yoshi, V. Kumar

Recursive Blocked Data Formats and BLAS's for Dense Linear Algebra Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

F. G u s t a v s o n , A. Henriksson, L Yonsson, B. KdgstrSm, P. Ling

Superscalar GEMM-based Level 3 BLAS - The On-going Evolution of a Portable and High-Performance Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

F. Gustavson, A. Henriksson, L Jonsson, B. KdgstrSm, P. Ling

Parallel Solution of Some Large-Scale Eigenvalue Problems Arising in Chemistry and Physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

D.L. Harrar II, M.R. Osborne

An Embarrassingly Parallel ab initio MD Method for Liquids . . . . . . . . . . . . 224 F. Hedman, A. Laaksonen

A New Parallel Preconditioner for the Euler Equations . . . . . . . . . . . . . . . . . 230 L. Hemmingsson, A. Kiihiiri

Parti t ioning Sparse Rectangular Matrices for Parallel Computat ions of Ax and ATv . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

B. Hendrickson, T.G. Kolda

NetLink: A Modern Data Distribution Approach Applied to Transparent Access of High Performance Software Libraries . . . . . . . . . . . . . . . . . . . . . . . . 248

L Holmqvist, E. LindstrSm

Modernization of Legacy Application Software . . . . . . . . . . . . . . . . . . . . . . . . . 255 J. Howe, S.B. Baden, T. Grimmett, K. Nomura

Parallel Methods for Fluid-Structure Interaction . . . . . . . . . . . . . . . . . . . . . . . 263 C . B . J e n s s e n , T. Kvamsdal, K.M. Okstad, J. Amundsen

Parallel Computing Tests on Large-Scale Convex Optimizat ion . . . . . . . . . . 275 M. Kallio, S. Salo

Parallel Sparse Matrix Computat ions in the Industrial Strength P I N E A P L Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

A.R. Krommer

Massively Parallel Linear Stability Analysis with P_ARPACK for 3D Fluid Flow Modeled with MPSalsa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286

R . B . L e h o u c q , A.G. Salinger

•

Parallel Molecular Dynamics Simulations of Biomolecular Systems . . . . . . . 296 A. Lyubartsev, A. Laaksonen

A Parallel Solver for Animal Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 P. Madsen, M. Larsen

Scheduling of a Parallel Workload: Implementat ion and Use of the Argonne Easy Scheduler at PDC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

L. Malinowsky, P. Oster

An Algorithm to Evaluate Spectral Densities of High-Dimensional Stat ionary Diffusion Stochastic Processes with Non-linear Coefficients: The General Scheme and Issues on Implementat ion with PVM . . . . . . . . . . . 315

Y.V. Mamontov, M. Willander

High-Performance Simulation of Evolutionary Aspects of Epidemics . . . . . 322 W. Maniatty, B.K. Szymanski, T. Caraco

A Parallel Algorithm for Computing the Extremal Eigenvalues of Very Large Sparse Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332

F. Manne

Technologies for Teracomputing: A European Option . . . . . . . . . . . . . . . . . . . 337 A. Mathis

High Performance Fortran: Status and Prospects . . . . . . . . . . . . . . . . . . . . . . . 345 P. Mehrotra, J. Van Rosendale, H . Z i m a

PAVOR - Parallel Adaptive Volume Rendering System . . . . . . . . . . . . . . . . . 357 M. Meiflner

Simulation Steering with SCIRun in a Distributed Environment . . . . . . . . . 366 M. Miller, C.D. Hansen,, C . R . J o h n s o n

Addressing the Requirements of ASCI-class Systems . . . . . . . . . . . . . . . . . . . . 377 J . H . M i r z a

A Parallel Genetic Algorithm for the Graphs Mapping Problem . . . . . . . . . 379 O.G. Monakhov, E.B. Grosbein

Parallel Wavelet Transforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 O. Mr Nielsen

Writing a Multigrid Solver Using Co-array Fortran . . . . . . . . . . . . . . . . . . . . 390 R.W. Numrich, J. Reid, K. Kim

Exploiting Visualization and Direct Manipulation to Make Parallel Tools More Communicative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400

C . M . P a n c a k e

•

Deploying Fault-Tolerance and Task Migration with NetSolve . . . . . . . . . . . . 418 J.S. Plank, H. Casanova, M. Beck, J . D o n g a r r a

Comparison of Implicit and Explicit Parallel Programming Models for a Finite Element Simulation Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

J. Ptaiek, K. Banal, J. Kitowski

Parallel Algorithms for Triangular Sylvester Equations: Design, Scheduling and Scalability Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

P. Poromaa

Fast and Quanti tat ive Analysis of 4D Cardiac Images Using a SMP Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447

V. Positano, M.F. SantareUi, L. Landini, A. Benassi

Ab Initio Electronic Structure Methods in Parallel Computers . . . . . . . . . . . 452 S. P S y k k 5

Iterative Solution of Dense Linear Systems Arising from Integral Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460

J. Rahola

Comparison of Parti t ioning Strategies for PDE Solvers on Multiblock Grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468

J. Rantakokko

Ship Design Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476 C. Risager, J . W . P e r r a m

Parallelization Strategies for the VMEC Program . . . . . . . . . . . . . . . . . . . . . . 483 L.F. Romero, E.M. Ortigosa , E.L. Zapata, J.A. Jirngnez

Rational Krylov Algorithms for Eigenvalue Computa t ion and Model Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491

A. R u h e , D. Skoogh

Solution of Distributed Sparse Linear Systems Using PSPARSLIB . . . . . . . . 503 Y. Saad, M. Sosonkina

Parallelization of the DAO Atmospheric General Circulation Model . . . . . . 510 W. Sawyer, R. Lucchesi, P. Lyster, L. Takacs, J. Larson, A. Molod, S. Nebuda, C. Pabon-Ortiz

Dynamic Performance Callstack Sampling: Merging TAU and DAQV . . . . . 515 S. Shende, A.D. Malony, S.T. Hackstadt

A Parallel Rational Krylov Algorithm for Eigenvalue Computat ions . . . . . . 521 D. Skoogh

xIv

Por table Implementa t ion of Real-Time Signal Processing Benchmarks on HPC Platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527

J. Suh, V.K. Prasanna

Large Scale Active Networks Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537 K. Swaminathan., R. Radhakrishnan, P.A. Wilsey, P. Alexander

Forward Dependence Folding as a Method of Communicat ion Opt imizat ion in SPMD Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543

Z. Szczerbinski

A Parallel Genetic Clustering for Inverse Problems . . . . . . . . . . . . . . . . . . . . . 551 H. Telega, R. Schaefer, E. Cabib

A Parallel Hierarchical Solver for Fini te Element Applicat ions . . . . . . . . . . . 557 C.-A. Thole, A. Supalov, S. Mayer

Parallel Computa t ion and Visualization of 3D, Time-Dependent , Thermal Convective Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565

P. Wang, P. Li

Recursive Formulat ion of Cholesky Algori thm in For t ran 90 . . . . . . . . . . . . . 574 J. Wagniewski, B.S. Andersen, F. Gustavson

High Performance Linear Algebra Package for FORTRAN 90 . . . . . . . . . . . . 579 J. Wa~niewski, Y. Dongarra

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585