Complex Systems Research on Parallel Computers

Complex Systems research is a multidisciplinary eld covering the gamut from Physics, Mathematics, Computer Science, Biology to Economics. Whilst there is no good agreement on what the term \Complex-ity" might mean, there are a number of threads or concepts linking the diierent disciplines together into a coherent eld of research. Traditional mathematical techniques such as Analysis, Linear Algebra and Diieren-tial Equations have been largely ineeective at tackling the problems in Complex Systems; it is only in the last twenty years or so that computer modeling has been able to shed light in this eld of research. The eld is maturing, and the complex systems models are getting more sophisticated. It is now becoming apparent that high performance computing and in particular parallel computing will need to be exploited in the next generation of models. Unfortunately, the structure of typical complex systems models have acted as an impediment to eective expoita-tion of their inherent parallelism. This paper discusses some of the issues, and suggests a way forward. The idea of what is a complex system is not well deened, however people intuitively feel that a biological organism is complex, whereas a cup of water is not. There are a number of themes which run through studies of complex systems , one of the more important being that of emergence. Whilst the notion of emergence is as notoriously diicult to pin down rigourously as the concept 1