SuperSim: Extensible Flit-Level Simulation of Large-Scale Interconnection Networks

The interconnection networks of modern largescale computing systems are quickly increasing in size and complexity to keep up with the demand for computing capability. These systems rely heavily on complex router microarchitectures and intelligent adaptive routing algorithms structured for cost-optimized low-diameter networks. These technologies need to be properly modeled and evaluated during design space exploration and for performance characterization of the system. We present SuperSim, an open-source flit-level interconnection network simulator that enables focused evaluation of issues related to designing and deploying large-scale highperformance networks. SuperSim is a programmer-centric simulation framework explicitly designed to be flexibly extended and is supported by a number of tools making it easy to use and allowing users to model systems quickly. In this work we show the results for simulation case studies demonstrating the power of SuperSim to uncover otherwise overlooked details in large-scale interconnection networks.

I. Introduction

For high-performance computing systems the interconnection network is the critical piece of hardware that makes the system a “supercomputer”. In other realms, cloud and enterprise data centers and personal computers for example, one can find the same processors, memory, storage devices, and accelerators. It is the network that tightly couples these devices in such a way that it can be viewed and used as a single high-performance system. The network enables the programming runtimes (e.g., MPI, SHMEM, UPC, etc.) to be highly productive to programmers by supplying large amounts of bandwidth and low message latencies. Supercomputer architectures are seeing a large increase in size and complexity to achieve an exaflop of performance [32]. As the number and complexity of nodes continues to increase to achieve exascale, the importance and critical role of the network also increases.