Fast event-based epidemiological simulations on national scales

TL;DR

This paper introduces a parallel computational framework for large-scale, data-driven epidemiological simulations that efficiently models infectious disease spread across national populations using shared memory architectures.

Contribution

It presents a novel parallel algorithm combining domain decomposition and dependency-aware scheduling for scalable, high-resolution epidemiological simulations on multi-core systems.

Findings

Scales efficiently on all physical cores with realistic workloads.

Resilient to various model configurations.

Enables solving inverse problems at national scale.

Abstract

We present a computational modeling framework for data-driven simulations and analysis of infectious disease spread in large populations. For the purpose of efficient simulations, we devise a parallel solution algorithm targeting multi-socket shared memory architectures. The model integrates infectious dynamics as continuous-time Markov chains and available data such as animal movements or aging are incorporated as externally defined events. To bring out parallelism and accelerate the computations, we decompose the spatial domain and optimize cross-boundary communication using dependency-aware task scheduling. Using registered livestock data at a high spatio-temporal resolution, we demonstrate that our approach not only is resilient to varying model configurations, but also scales on all physical cores at realistic work loads. Finally, we show that these very features enable the solution of inverse problems on national scales.