FLARES: A Framework for Large-Scale Agent-Based Rapid Epidemic Simulation

Agent-based modeling (ABM) is a powerful simulation methodology employed to analyze complex systems by representing individual entities, known as agents, which interact autonomously within a defined set of rules. This approach is particularly effective for epidemic simulation, where capturing the nuanced interactions and behaviors of individuals is crucial for understanding disease dynamics and spread. However ABM often come with complected and slow implementation. To address the performance issue, in this paper, we introduce a Framework for Large-scale Agent-based Rapid Epidemic Simulation (FLARES), a novel GPU-accelerated framework designed to enhance the performance of agent-based infectious disease transmission model. FLARES provides robust support for parallel execution, significantly reducing the overall time consumption for high computational cost tasks, enables researchers to conduct detailed and accurate epidemic analysis of disease spread.