Title: JiST: Virtual machine-based simulation Speaker: Rimon Barr Date: TBD Time: TBD Location: TBD From physics to biology, from weather and stock predictions to estimating the performance of a new processor design, people in many avenues of science and industry increasingly depend on software simulation. Research in the areas of wireless ad hoc and sensor networks, for example, is fundamentally dependent on simulators. Yet, existing tools for modeling the behaviour of such networks are often unsatisfactory, because they severely limit the possible scale, level of detail or duration of the simulation results. Discrete event simulators have been the subject of much research. Systems researchers have built many types of simulation kernels and libraries, while the languages community has designed numerous languages specifically for simulation. In this talk, I will describe a new approach for constructing discrete event simulators that leverages virtual machines and combines the traditional systems-based and language-based approaches to simulator construction. JiST, for Java in Simulation Time, is a general-purpose simulation engine that embodies this new technique. It embeds simulation execution semantics directly into the Java virtual machine. The system provides all the standard benefits that the modern Java runtime affords. In addition, JiST is efficient, out-performing existing highly optimized simulation runtimes, and inherently flexible, capable of transparently performing cross-cutting program transformations and optimizations. I will explain how the basic system works and discuss various extensions. Finally, I will illustrate a practical application of JiST through SWANS, a Scalable Wireless Ad hoc Network Simulator that runs atop JiST. SWANS can simulate million node wireless networks, which represents two orders of magnitude increase in scale over what existing simulators can achieve on equivalent hardware and at the same level of detail.