High performance computing challenges in social simulation

Session chairs:

Bogumił Kamiński –

Przemysław Szufel –



We invite submissions both on methodology and example implementations of complex, large-scale agent-based models. Of particular interest are submissions presenting the following topics:

  • Agent based models incorporating full synthetic populations of entire communities (1:1 scale simulation models).
  • Simulation models using multiple modeling methods, in particular, but not limited to: artificial intelligence, optimization, differential equations, computer algebra.
  • Models having scale that requires coping with optimal computing allocation methods.
  • Simulations implemented and run in distributed computing architectures including cloud and supercomputing environments.
  • Large scale simulation experiment design, execution, and analysis.
  • Technical problems in massive scale execution of simulation such as performance, out-of-core computing, data flow, grid computing issues.

The track is accompanied by the workshop “An introduction to agents-based simulations in Julia language” that is organized during the conference. It is planned to be divided into four 1-hour sessions:

  • Part one: First steps with the Julia language (introduction to the language aimed at beginners); instructor: Bogumił Kamiński
  • Part two: Meet Agents.jl (learning the basics of development of agent-based models using Agents.jl in the Julia language); instructors: George Datseris, Timothy DuBois
  • Part three: Use Agents.jl (interactive session: development of a sample agent-based model using Agents.jl by the participants); instructors: George Datseris, Timothy DuBois
  • Part four: Agent-based modeling at scale (presentation of example large scale agent-based models created using the Julia language); instructor: Przemysław Szufel

Anyone interested is welcome to participate in it.

You might want to read the following paper before the workshop to get the flavor of what Agents.jl has to offer: Agents.jl: A performant and feature-full agent based modelling software of minimal code complexity, George Datseris, Ali R. Vahdati, Timothy C. DuBois, 2021,