ROSS: A Massively Parallel Discrete-Event Simulator for Modeling Extreme-Scale Computer Systems
Abstract:
Rensselaer's Optimistic Simulation System is a framework for developing large-scale event-driven models that are able to execute on the world's largest supercomputers. In this talk, I will first present an overview of discrete-event simulation and PDES followed by ROSS' design and implementation. Benchmark performance results are presented next where super-linear performance is shown when executing on nearly 2,000,000 processor cores. The secret to ROSS' performance is the use of "reverse computation" to efficiently support the optimistic processing of model events.
Two key applications that use ROSS are then presented. The first is CODES which is an extreme-scale network and storage system modeling framework. CODES enables models of a number of HPC network topologies including dragonfly, fat tree, slim fly and torus consisting of 1000's, millions and even a billion network node configurations. The second application is NeMo which is a spiking neuromorphic processor simulator which is able to model processor designs with over 8 million neuromorphic processing cores where each core has 256 neurons and 4096 synapses.
Bio:
Professor Chris Carothers is a faculty member in the Computer Science Department at Rensselaer Polytechnic Institute. He received the Ph.D., M.S., and B.S. from Georgia Institute of Technology in 1997, 1996, and 1991, respectively. Prior to joining RPI in 1998, he was a research scientist at the Georgia Institute of Technology. His research interests are focused on massively parallel computing which involve the creation of high-fidelity models of extreme-scale networks and computer systems. These models have executed using nearly 2,000,000 processing cores on the largest leadership class supercomputers in the world. Additionally, Professor Carothers serves as the Director for the Rensselaer Center for Computational Innovations (CCI). CCI is a partnership between Rensselaer and IBM. The center provides computational and storage resources to a diverse network of researchers, faculty, and students from Renssleaer, government laboratories, and companies across a number of science and engineering disciplines. The flagship supercomputer is an 8 peta-flop (PF) IBM AC922 hybrid CPU/GPU supercomputer named "AiMOS".