- Copyright © 2002 Society of Exploration Geophysicists
The session dedicated to high-performance computing and visualization in medical, geoscience, and remote sensing included seven oral presentations and two demos. The session was very strongly weighted toward visualization with but one paper on high-performance computing.
The presentation of keynote speaker Linda Fellingham (from the FastApps group at Sun Microsystems), “Medical and Earth Resources Examples of HPC Visualization” gave a perspective on visualization based on more than 20 years of experience. In her view, high-performance computing, using larger numbers of processors and massive amounts of memory, is creating increasingly larger and more complex data sets, and interactive visualization is becoming more important for the analysis and understanding of these massive data sets.
Traditionally, the problem has been split into two parts—a computational back-end and visualization front-end. This approach has significant limitations. The user must at times split the application between a compute server and a visualization system. This structure and the requirement to transfer data between platforms impose artificial limitations on the interactivity of the visualization and data analysis. It also limits interaction between visualization and computation, particularly when early visualization of computation results might advantageously alter the model or direct the computation. To demonstrate the value of integrated high-performance computing and visualization, Fellingham presented interactive surgical simulation programs in current use at Stanford University and Harvard/Brigham & Women's University.
Large-scale visualization environments
Mary Cole of MechDyne, and formerly of ARCO's Visualization Technology group, discussed “Large Scale Visualization Environments—Issues and Options.” She described various large-scale environments and the trade-offs in performance, function, and application between different large-screen systems. Her observations are based on more than 15 years of experience in medical visualization and geoscience visualization.
Large-scale visualization environments allow groups to collaboratively explore, interact with, and modify data sets. Over the past 5–10 years, these systems have been applied to a wide range …