Visualization in Supercomputing

Visualization in Supercomputing

Raul H. Mendez Editor Visualization in Supercomputing With 166 Illustrations, 25 in Color Springer-Verlag New York Berlin Heidelberg London Paris Tokyo Hong Kong

Raul H. Mendez Institute for Supercomputing Research 15F Inui Building Kachidoki 1-31-1 Kachidoki, Chuo-ku Tokyo 104 Japan About the cover: The object is the quaternion Julia set for the function $z 2 + 0.2809 + 0.53i$. The upper front left octave is missing, revealing the basins of attraction of the four-cycle defining its interior. These interior points are such that they remain bounded under repeated applications of the function. In fact, they are periodic, moving from red to green to yellow to blue to red again. The image was ray-traced on an AT&T Pixel Machine 964dX by John C. Hart at the Electronic Visualization Laboratory at the University of Illinois at Chicago. Copyright 1990 John C. Hart. Library of Congress Cataloging-in-Publication Data Visualization in supercomputing/editor, Raul H. Mendez. p. cm. ISBN-13:978-1-4612-7971-6 I. Supercomputers. I. Mendez, R. (Raul) QA76.5.V55 1990 004. 1 'l-dc20 89-48589 Printed on acid-free paper. 1990 Springer-Verlag New York, Inc. Softcover reprint of the hardcover 1st edition 1990 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer-Verlag New York, Inc., 175 Fifth Avenue, New York, NY 10010, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use of general descriptive names, trade names, trademarks, etc., in this publication, even if the former are not especially identified, is not to be taken as a sign that such names, as understood by the Trade Marks and Merchandise Marks Act, may accordingly be used freely by anyone. Camera-ready copy prepared by authors. 987654321 ISBN-13:978-1-4612-7971-6 DOl: 10.1007/978-1-4612-3338-1 e-isbn-13:978-1-4612-3338-1

Preface In August 1988, one of the finest groups of supercomputing and visualization researchers from the U.S. and Japan met In Tokyo for a four-day conference. This conference was unique in that it was the first gathering of the countries' leading scientists in computer visualization and high-speed computing. It was also one of the rare opportunities for visualization scientists and hardward designers to discuss and exchange views in this excitinq; field. The idea of visualization IS not new; because massive amounts of numeric data are far more comprehensible when converted into graphical form, visualization is becoming an intimate part of many areas of research. Techniques for visualization, however, are still being developed, and visualization research is just beginning to be recognized as a cornerstone of future computational science. As scientists bring increasingly complex problems to computational science, visualization will become an even more essential tool for extracting science out of numbers. The goal of this conference was to provide a focal point for users and hardward/software designers, scientists from both ends of the visualization spectrum, to share their progress and discuss their needs. The conference site itself, one of the world's centers of computer technology, provided an auspicious environment for a successful conference. The present volume is a collection of papers selected from those presented at the conference. It is divided into three parts, representing the major areas in visualization computing: visualization applications, hardware/performance, and visualization theory. In Part 1, Visualization Applications, Kozo Fujii's paper reviews visualization methods used in computational fluid dynamics research. In particular, he discusses the hardware requirements for doing effective viusalization of fluid dynamics from the researcher's point of view. The time-to-solution, an often overlooked aspect of visualization, is addressed in Kumazawa's paper. In this article, various system configurations are examined to study the turnaround time from computer simulation to grahpical display. Having the fastest computer does not necessarily mean the shortest time-to-solution. Part 2, covering visualization hardware and performance, includes a paper by Robert Fulton and Kuo-Ning Chiang on the use of parallel/vector computers with finite-element systems. The paper by Kok-Meng Lue and Kazuto Miyai addresses the basic computational performance of two graphics supercomputers, the Ardent Titan and Stellar GS1000. In Part 3, Alvy Ray Smith's paper covers important topics such as volume and 3D surface imaging, geometry-based ~ r a p hworkstations i c s vs. image computers. The paper also discusses the applicability of the volume imaging concept in various fields. Lastly, I have collected in the appendix viewgraphs from several presentations at the conference which should complement the materials presented. Raul H. Mendez Tokyo, March 1989

Contents Preface... v Part 1: Visualization Applications Supercomputers and Workstations in Fluid Dynamics Research... 3 Kozo Fujii Numerical Simulation of a 3-D Backward-Facing Step Flow... 14 Hiroshi Takeda and Erika Misaki System Solutions for Visualization: A Case Study... 25 Kohei Kumazawa and Christopher Eoyang Part 2: Visualization Hardware/Performance A General Approach to Nonlinear Dynamic Analysis on ParallellVector Computers...................................................... 41 Robert E. Fulton and Kuo-Ning Chiang Basic Performance of Two Graphics Supercomputers: Stellar GS 1000 and Ardent Titan-2... :... 64 Kok-Meng Lue and Kazuto Miyai High Bandwidth Interactivity and Super Networks... 80 James Perdue Cellular Array Processor CAP and Visualization... 100 Hiroyuki Sato, Mitsuo Ishii, Morio Ikesaka, Kouichi Murakami, and H iroaki I shihata Requirements for Scientific Visualization: Evolution of an Accelerator Architecture.................................................. 117 Mary C. Whitton Part 3: Visualization Theory Advanced Visualization Environments: Knowledge-Based Image Modeling.............................................................................. 135 Bruce H. McCormick Geometry vs Imaging: Extended Abstract... 151 Alvy Ray Smith Lighting Simulation... 157 Eihachiro Nakamae Appendix Co-processing Environments for Interactive Visualization....... 169 Craig Upson Supercomputing Environments for the 1990s... 184 Larry Smarr