Parallel Numerical Algorithms
in Computational Science and Engineering
Ahmed H. Sameh
Department of Computer Science
Purdue University
In this presentation, we provide a definition of the interdisciplinary
field of Computational Science and Engineering (CSE) and point out its
pivotal role in advancing certain science and engineering
disciplines. Such advances, however, require extensive large-scale and
large-scope numerical simulations. These, in turn, can only be
realized through the use of high-end parallel computing platforms. One
often discovers, however, that simple implementation of the key
classical sequential algorithms on these novel architectures leads to
poor overall performance. High performance is thus most frequently
realized through the redesign of algorithms that take maximum
advantage of the modern parallel architectures in which floating point
operations are the least time-consuming tasks. We provide few
numerical linear algebra examples to illustrate this point with
applications in computational mechanics and nanoelectronics.
Brief Biography:
Ahmed H. Sameh is the Samuel D. Conte Professor of Computer
Science. His current research interests include numerical linear
algebra, and Parallel Computing. He joined Purdue in 1997 as Head of
Computer Science, after being Head of Computer Science at the
University of Minnesota, Minneapolis, and the holder of the William
Norris Chair for Large-Scale Computing. He was a faculty member in the
Department of Computer Science at the University of Illinois at
Urbana-Champaign, from 1968 to 1992. He is a Fellow of ACM, IEEE, and
AAAS, and a member of SIAM. He has also received the IEEE's 1999 Harry
Goode Award for "seminal and influential work in parallel numerical
algorithms."