Performance Considerations of Shared Virtual Memory Machines

Generalized speedup is deened as parallel speed over sequential speed. In this paper the generalized speedup and its relation with other existing performance metrics, such as traditional speedup, eeciency, scalability, etc., are carefully studied. In terms of the introduced asymptotic speed, we show that the diierence between the generalized speedup and the traditional speedup lies in the deenition of the eeciency of unipro-cessor processing, which is a very important issue in shared virtual memory machines. A scientiic application has been implemented on a KSR-1 parallel computer. Experimental and theoretical results show that the generalized speedup is distinct from the traditional speedup and provides a more reasonable measurement. In the study of diier-ent speedups, an interesting relation between xed-time and memory-bounded speedup is revealed. Various causes of superlinear speedup are also presented.