Visualization In Supercomputing

Visualization in Supercomputing

Massive amounts of numeric data are far more comprehensible when converted into graphical form. Hence visualization is becoming an integral part of many areas of research. The idea of visualization is not new, but techniques for visualization are still being developed, and visualization research is just beginning to be recognized as a cornerstone of future computer science. As scientists handle increasingly complex problems with computers, visualization will become an even more essential tool for extracting sense from numbers. This volume is a collection of the best papers selected from those presented at the August 1988 Visualization in Supercomputing Conference in Tokyo, Japan. It is divided into three parts: visualization applications, hardware and performance, and visualization theory. Subjects covered include visualization methods used in computational fluid dynamics research, time-to-solution aspects of visualization, the use of parallel/vector computers with finite element method systems, basic computational performance of two graphics supercomputers, and the applicability of the volume imaging concept in various fields.

Visualization in Supercomputing

Visualization in Scientific Computing

Visualization in scientific computing is getting more and more attention from many people. Especially in relation with the fast increase of com puting power, graphic tools are required in many cases for interpreting and presenting the results of various simulations, or for analyzing physical phenomena. The Eurographics Working Group on Visualization in Scientific Com puting has therefore organized a first workshop at Electricite de France (Clamart) in cooperation with ONERA (Chatillon). A wide range of pa pers were selected in order to cover most of the topics of interest for the members of the group, for this first edition, and 26 of them were presented in two days. Subsequently 18 papers were selected for this volume. 1'he presentations were organized in eight small sessions, in addition to discussions in small subgroups. The first two sessions were dedicated to the specific needs for visualization in computational sciences: the need for graphics support in large computing centres and high performance net works, needs of research and education in universities and academic cen tres, and the need for effective and efficient ways of integrating numerical computations or experimental data and graphics. Three of those papers are in Part I of this book. The third session discussed the importance and difficulties of using stan dards in visualization software, and was related to the fourth session where some reference models and distributed graphics systems were discussed. Part II has five papers from these sessions.