Argonne National Laboratory

Barry Smith named Argonne Distinguished Fellow

July 10, 2015

Barry Smith, a senior computational mathematician in Argonne’s Mathematics and Computer Science Division, has been named an Argonne Distinguished Fellow, the laboratory’s highest scientific and engineering rank.

The Argonne Distinguished Fellow title is comparable in stature to an endowed chair at a top-ranked university and recognizes exceptional contributions in a person's field. The rank is given for sustained outstanding scientific and engineering research and can also be associated with outstanding technical leadership of major, complex, high-priority projects.

Barry received his bachelor’s degree in mathematics from Yale University in 1986 and his Ph.D. in mathematics from the Courant Institute of Mathematical Science (New York University) in 1990. He held the Wilkinson postdoctoral fellowship at Argonne from 1990 to 1992 and then was an assistant professor at the University of California at Los Angeles for two years. In 1994, Barry rejoined Argonne, where he soon became deeply involved in the development of PETSc, the Portable Extensible Toolkit for Scientific computing. Barry’s work has transformed how large-scale software libraries are developed, supported, and used – not only within DOE, but also across the wider scientific and engineering community.

Barry has also worked extensively with applications scientists, devising methods for the solution of computational science problems ranging from traditional areas such as fusion simulation, ice sheet modeling, and thermohydraulics to emerging fields such as computational biology. In 2010 Barry (with his colleague Lois McInnes) won the prestigious E.O. Lawrence Award for outstanding contributions in research and development supporting the Department of Energy and its missions; and in 2015 the core PETSc development team led by Barry won the SIAM/ACM Prize in Computational Science and Engineering, a prize awarded only every two years.

Barry’s role as a leader in the scientific computing community extends far beyond his development of PETSc. Indeed, Barry pioneered the development of highly parallel domain decomposition methods for linear elasticity problems in three dimensions. His innovative approaches resulted in algorithms whose convergence was independent of the grid spacing, problem size, and number of processors. In recognition of these achievements, he won second prize at the 1991 Fifth Leslie Fox Prize meeting, an international award in numerical analysis offered every two years, and he was co-winner of the 1993 Householder Prize for best dissertation in numerical linear algebra during the previous three years.

Barry’s recent work has focused on designing and implementing efficient solvers for cutting-edge simulations on DOE leadership-class computers and other advanced architectures. For example, Barry is project contact for the new FASTMath SciDAC Institute, which is developing and deploying scalable mathematical algorithms and software tools for reliable simulation of complex physical phenomena. This project addresses one of the key challenges facing the scientific computing community: the shift to multicore/many-core nodes and million-way parallelism. Barry is also tackling the challenges that are looming with the emerging exascale era. His work on the IDEAS (Interoperable Design of Extreme-scale Application Software) project, for example, is helping address memory bandwidth limitations in sparse matrix computations, a severe bottleneck in many DOE simulations.