Abstract: A performance comparison between an optimistic and a conservative parallel simulation kernel and a sequential kernel is presented. A spectrum of ATM network and traffic scenarios representative of those used by ATM networking researchers are used for the comparison. Experiments are conducted using a cell-level ATM network simulator, and an 18-processor SGI Power Challenge shared-memory multiprocessor.
The results show the performance advantages of parallel simulation over sequential simulation for ATM networks. Absolute speedups of 4-5 are attained on 16 processors on several ATM benchmark scenarios and the optimistic kernel achieves 2 to 5 times speedup on all seven benchmarks. However, the relative performance of the two parallel simulation kernels is dependent on the size of the ATM network, the number of traffic sources, and the traffic source types used in the simulation. For this type of problem the best single point performance is provided by the conservative kernel even on a single processor. Unfortunately, the conservative performance is susceptible to small changes in the modeling code and it outperformed by the optimistic kernel on 5 of the 7 benchmarks. The optimistic parallel simulation kernel thus provides more robust performance but its absolute speedup is limited by the overheads of its implementation which make it approximately half the speed of the sequential kernel on one processor.
These performance results represent the first comparative analysis of parallel simulation for a spectrum of realistic, irregular, low granularity, communication network models.
Keywords: Parallel simulation, parallel performance, ATM network modeling, TimeWarp, conservative synchronization, optimistic synchronization. conservative simulation