For the performance study, we use two applications: (i) a 3-level cascading network of Asynchronous Transfer Mode (ATM) multiplexors [86] and (ii) a simple Personal Communications Services (PCS) network using fixed channel assignment [16,18,19]. We observe as much as four and six times improvement in execution speed over traditional state-saving for the respective applications [20,21]. In terms of overall execution rates, we have observed PCS model event rates as high as 800,000 and speedups of 15 on a 16 processor SGI Origin 2000 [64].