Chapter 4

An Overview of the SimWorld Agent-Based Grid Experimentation System

Matthias Scheutz

Department of Computer Science, Tufts University, Medford, MA, USA

Jack J. Harris

Human Robot Interaction Laboratory, Indiana University, Bloomington, IN, USA


Computational modeling is becoming increasingly important, even in fields that have not traditionally used computational models (e.g., archaeology or anthropology). Researchers in both the natural and social sciences employ computer simulations to elucidate the time course of physical and nonphysical processes, or to explore the dynamics among different interacting entities, in an effort to discover new relationships that might lead to generalizable laws or to verify hypothesized principles as part of the empirical discovery loop (Peschl and Scheutz, 2001). However, there are two main obstacles to making effective use of today’s (and likely also tomorrow’s) computing environments. First, navigating the complexity associated with running large-scale computational simulations requires detailed knowledge about the available high-performance computing (HPC) environments. Such prerequisite knowledge includes how to set up a simulation on the host computers (possibly including compilation on the target platform with installation of all the required libraries), how to schedule sets of simulations through the batch system, how to retrieve the resultant data, and how to troubleshoot if simulations do not finish (because they ...

Get Large-Scale Computing Techniques for Complex System Simulations now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.