The quest for scientific knowledge has been aided by the continued growth of computational power. This growth has led to unprecedented amounts of scientific data that require analysis and visualization to extract scientific insight. In addition to the complexities of large data, the systems used for scientific inquiry are complex, heterogenous, distributed, workflow-driven and produce results with differing temporal and spatial fidelity. This talk will motivate these challenges with large-scale scientific campaigns performed on the largest supercomputing systems in the world and discuss our research and development efforts to address these challenges in three different areas: 1) cost models for in situ processing, 2) visualization on heterogenous hardware, and 3) new paradigms for scientific visualization.

bio: Dave Pugmire is a Distinguished Research Scientist and Visualization Group Leader at the Oak Ridge National Laboratory. He also holds a Joint Faculty Professor appointment in the Electrical Engineering and Computer Science Department at the University of Tennessee. Prior to coming to Oak Ridge he was a research scientist at Los Alamos National Laboratory. He received his Ph.D. in 1999 from the University of Utah. He is interested in large-scale visualization, scalable algorithms and in situ processing. He’s has played a leading role in the development of several visualization products, including EAVL, VTK-m, VisIt and FIDES.