Adaptive Parallelism with Piranha

\Adaptive parallelism" refers to parallel computations on a dynamically changing set of processors: processors may join or withdraw from the computation as it proceeds. Networks of fast workstations are the most important setting for adaptive parallelism at present. Workstations at most sites are typically idle for signiicant fractions of the day, and those idle cycles may constitute in the aggregate a powerful computing resource. For this reason and others, we believe that adaptive parallelism is assured of playing an increasingly prominent role in parallel applications development over the next decade. The \Piranha" system now up and running on a heterogeneous network at Yale is a general-purpose adaptive parallelism environment. It has been used to run a variety of production applications, including applications in graphics, theoretical physics, electrical engineering and computational uid dynamics. In this paper we describe the Piranha model and several archetypal Piranha algorithms. Our main goals are to show that it is an eeective and workable approach, to explain how the Piranha model constrains algorithm design, and to suggest directions for theoretical and practical research.