Title:  Devising Effective Parallel Algorithms

Abstract:

The field of parallel computing seems to be in a continual state
of flux with rapid evolution in architectures, programming models
and languages.  The one constant is the overriding importance of
good algorithms.  Unfortunately, devising effective algorithms for
scientific applications remains more of an art than a science.
In this talk I will describe new parallel algorithms for two
important classes of scientific computations: simulating the
motions of molecules, and modeling car crashes.  These two,
very different, examples will be used as case studies to address
broader questions concerning tools, techniques and philosophy
underlying the development of effective parallel algorithms and
applications.