Proceedings 15th Workshop on Parallel and Distributed Simulation (2001)
Lake Arrowhead, California
May 15, 2001 to May 18, 2001
James Nutaro , University of Arizona
Hessam Sarjoughian , University of Arizona
Optimistic and conservative simulation algorithms have been effective for speeding up the execution of many simulation programs. Event stepped techniques have also been shown to be effective for certain types of problems. This paper presents a conjecture that sparse systems are effectively simulated by conservative and optimistic techniques, where sparseness is characterized by the ratio of output generating states to non-output generating states. The speedup results for a representative sparse simulation are shown to support this conjecture.
J. Nutaro and H. Sarjoughian, "Speedup of a Sparse System Simulation," Proceedings 15th Workshop on Parallel and Distributed Simulation(PADS), Lake Arrowhead, California, 2001, pp. 193.