Time Warp and Breathing Time Buckets are two general-purpose optimistic synchronization strategies for supporting parallel discrete-event simulations. However, each one of these approaches has potential fatal shortcomings. Time Warp may exhibit rollback explosions that can cause an avalance of antimessages. Breathing Time Buckets, on the other hand, may not be able to process enough events per synchronization cycle to remain efficient. A new strategy, called Breathing Time Warp , has been developed in the Synchronous Parallel Environment for Emulation and Discrete-Event Simulation (SPEEDES) operating sytem. This new strategy solves both of these problems by mixing the two algorithms together, resulting in the best of both methods. This paper describes the implementation of the Breathing Time Warp algorithm in SPEEDES, and then shows how this new approach sometimes improves the performance of parallel discrete-event simulations.

No abstract

Time Warp and Breathing Time Buckets are two generalpnrpose optimistic synchronization strategies for supporting parallel discrete-event simulations. However, each one of these approaches has potential fatal shortcomings. Time Warp may exhibit rollback explosions that can cause an avalanche of antimessages. Breathing Time Buckets, on the other hand, may not be able to process enough events per synchronization cycle to remain efficient. A new strategy, called Breathing Time Warp, has been developed in the Synchronous Parallel Environment for Emulation and Discrete-Event Simulation (SPEEDES) operating system. This new strategy solves both of these problems by mixing the two algorithms together, resulting in the best of both methods. This paper describes the implementation of the Breathing Time Warp algorithm in SPEEDES, and then shows how this new approach sometimes improves the performance of parallel discrete-event simulations.

The event horizon is a very important concept that is useful for both parallel and sequential discrete-event simulations. By exploiting the event horizon, parallel simulations can process events in a manner that is risk-free (i.e., no antimessages) in adaptable “breathing” time cycles with variable time widths. Additionally, exploiting the event horizon can greatly reduce the event list management overhead that is common to virtually all discrete-event simulations. This paper develops an analytic model describing the event horizon from first principles using equilibrium considerations and the hold model (where each event, when consumed, generates a single new event with future-time statistics described by a known probability function). Exponential and Beta-density functions are used to verify the mathematics presented in this paper.

No abstract