Colloquium Details
Faculty Search Colloquium: Advancing the State of Scientific Programming
| Author: | Matthew Sottile |
|---|---|
| Date: | April 28, 2009 |
| Time: | 10:00 |
| Location: | 220 Deschutes |
| Hosts: |
Abstract
Computing has had a long-standing presence in the sciences. With increased capability to produce tremendous quantities of data through simulation and physical measurement, the level of sophistication required in computational activities is increasing. Supporting methods for expressing these activities must meet the demands created by this increase in sophistication. In particular, the question of how one expresses complex computations must be reexamined in scientific computing. How do current language techniques slow down scientific users seeking to perform science, and what can be done to make them more productive? The increase in scale of computational science investigations also means that issues in high performance computing must also be considered, ranging from parallelism on the desktop to traditional HPC platforms. This talk will lay out a description of an ongoing body of work investigating these areas and applying them to problems in the biological sciences.
Biography
Matthew Sottile is a research associate and adjunct instructor at the University of Oregon Department of Computer and Information Science. He received his B.S. in Mathematics and Computer Science in 1999 and M.S. in Computer Science in 2001 from the University of Oregon. He received his Ph.D. in Computer Engineering from the University of New Mexico in 2006.
Prior to joining the University, Dr. Sottile was a Technical Staff Member at the Los Alamos National Laboratory from 2001 through 2007 as a member of both the Advanced Computing Laboratory group and Continuum Dynamics group. His research work revolves around a common theme of improving the state of scientific computing to make it easier for scientists to perform complex activities. This research includes investigations in high performance computing, language and compiler development, and image processing algorithms. His work in high performance computing at LANL contributed to his team earning a 2005 R&D100 award and two National Nuclear Security Administration awards of excellence.
His current research is funded by the Department of Energy Office of Science and Placental Analytics, LLC.