Dynamically Adapting the Degree of Parallelism with Reflexive Programs

. In this paper we present a new method for achieving a higher cost--efficiency on parallel computers. We insert routines into a program which detect the amount of computational work without using problem-- specific parameters and adapt the number of used CPUs at runtime under given speedup/efficiency constraints. Several user--tunable strategies for selecting the number of processors are presented and compared. The modularity of this approach and its application--independence permit a general use on parallel computers with a scalable degree of parallelism. 1 Introduction Programs on parallel computers usually use all available processors. This is a waste of resources if the load is not evenly distributed or the amount of work is too small to justify a further partitioning. Our goal is to reduce the costs 1 by adapting the number of used processors dynamically according to the current load. In order to do so, the program observes its parallel routines and controls the degree of para..