Pose: Getting over grainsize in parallel discrete event simulation

Parallel discrete event simulations (PDES) encom-pass a broad range of analytical simulations. Their utility lies in their ability to model a system and pro-vide information about its behavior in a timely manner. Current PDES methods provide limited performance im-provements over sequential simulation. Many logical models for applications have fine granularity making them challenging to parallelize. In POSE, we exam-ine the overhead required for optimistically synchroniz-ing events. We have designed an object model based on the concept of virtualization and new adaptive op-timistic methods to improve the performance of fine-grained PDES applications. These novel approaches exploit the speculative nature of optimistic protocols to improve single-processor parallel over sequential per-formance and achieve scalability for previously hard-to-parallelize fine-grained simulations.1 1