Jim Stewart is the manager of the Optimization and Uncertainty Quantification Department at Sandia National Laboratories, a position he has held since 2007. The mission of his department is to provide leadership in the research, development, and application of scientific optimization and UQ algorithms and software. The department is well known for its Dakota optimization and UQ toolkit. Prior to this assignment, Jim was manager of the Advanced Computational Mechanics Architectures Department, where he led the growth of Sandia’s Sierra computational mechanics framework. His expertise spans the broad areas of computational mechanics, adaptive finite element methods, and verification and validation. Jim served on the Executive Council of the U.S. Association for Computational Mechanics from 2010-2014. He holds a B.S. and M.S. in Mechanical Engineering from University of Illinois, and a Ph.D. in Mechanical Engineering from Stanford University.
Science-Based Computational Modeling: New Questions, New Challenges, and Big Data Problems
3:00 p.m., Monday, March 28, 2016
1014 H.H. Dow (2300 Hayward St)
The state-of-the-art of computational modeling and simulation is anything but static. Tremendous progress has been achieved in the ability to model and predict complex behavior in multi-scale, multi-physics systems and materials. New questions can now be asked through our simulations, such as how sensitive are the quantities of interest to important input parameters, what is the uncertainty in the results, how credible are the answers, how can the models be improved, and how can simulations support critical design, economic, or safety decisions? Such progress is bound to continue, but it will require an expanded community of experimentalists, computer scientists, statisticians, and mathematicians. Moreover, the push to exascale computing is placing new demands on our algorithms, application software, programming models, and analysis workflows. Big data, and big data problems, are everywhere. This talk will delve into these questions, and discuss new research intended to address some of the challenges in the areas of uncertainty quantification, design optimization, and data assimilation.