With significant advances in computation, measurement, and storage technologies, terascale datasets have become increasingly commonplace. Numerical simulations running on supercomputing platforms and measurement techniques such as electromagnetic imaging can generate extremely large amounts of data. This data must be effectively interpreted in order to explain the underlying physical phenomena. The large volume of data associated can make this process very difficult. Conventional approaches relying on faster visualization techniques, data analysis, and use of parallelism, are likely to fail or have limited scope when used in isolation. A comprehensive end-t-end framework that integrates the data source, storage, servers, network, and the visualization client is critical for delivering scalable performance across a range of hardware platforms and datasets. We aim to develop such a framework for interrogative visualization of terascale datasets. The basis of this framework is a suite of compressed multiresolution representation and data streaming techniques that adapt in an error-controlled manner to available computational resources. Used in conjunction with fully threaded visualization servers and client ends, we attempt to provide seamless scalable performance. At the Computational Visualization Center at the University of Texas, we are engaged in a long-term project with the goal of developing a comprehensive framework for multi-scale visualization and simulation for terascale problems. This project deals strictly with progressive interrogative visualization of offline terascale datasets. Subsequently, we will extend the work to support interrogative steering of terascale physical simulations as well. The key terascale data analysis and visualization tasks will leverage from prior work on these problems for large data.

This material is based upon work supported by the National Science Foundation under Grant No. 9982297

Any options, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.