A Large-Scale Epidemic Simulation Framework for Realistic Social Contact Networks

Global pandemics can wreak havoc and lead to significant social, economic, and personal losses. Preventing the spread of infectious diseases requires implementing interventions at different levels of government, and evaluating the potential impact and efficacy of those preemptive measures. Agent-based modeling can be used for detailed studies of epidemic diffusion and possible interventions. We present Loimos, a highly parallel simulation of epidemic diffusion written on top of Charm++, an asynchronous task-based parallel runtime. Loimos uses a hybrid of time-stepping and discrete-event simulation to model disease spread. We demonstrate that our implementation of Loimos is able to scale to large core counts on an HPC system. In particular, Loimos is able to simulate a US-scale synthetic interaction network in an average of 1.497 seconds per simulation day when executed on 16 nodes on Rivanna at the University of Virginia, processing around 428 billion interactions (person-person edges) in under five minutes for an average of 1.4 billion traversed edges per second (TEPS).