AN APPROACH TO EXTREME-SCALE SIMULATION OF NOVEL ARCHITECTURES

We outline an approach for simulating computing architectures applicable to extreme-scale systems (thousands of processors) and to advanced, novel architectural configurations. We believe that simulation is the predictive tool of choice for evaluating the performance of such systems. Our component-based design allows for the seamless assembly of architectures from representations of workload, processor, network interface, switches, etc., with disparate resolutions into an integrated simulation model. This accommodates different case studies that may require different levels of fidelity in various parts of a system. Our initial prototype, comprising low-fidelity models of workload and network, aims to model at least 4096 computational nodes in a fat-tree network. It supports studies of simulation performance and scaling rather than the properties of the simulated system themselves. Future work will allow more realistic simulation and visualization of ASCI-like workloads on very large machines.