Parallel cellular programming for emergent computation

Abstract

The main goal of programming languages and tools has always been to make the programmer more productive and the programming task more effective. Appropriate programming languages and tools may drastically reduce the costs for building new applications as well as for maintaining existing ones. It is well known that programming languages can greatly increase programmer?s productivity by allowing the programmer to write high-scalable, generic, readable and maintainable code. Also, new domain specific languages, such as CA languages, can be used to enhance different aspects of software engineering. The development of these languages is itself a significant software engineering task, requiring a considerable investment of time and resources. Domain-specific languages have been used in various domains and the outcomes have clearly illustrated the advantages of domain specific-languages over general purpose languages in areas such as productivity, reliability, and flexibility. The main goal of the paper is answering the following question: How does one program emergent systems through cellular automata on parallel computers? We think that it is very important for an effective use of cellular automata for computational science on parallel machines to develop and use high-level programming languages and tools that are based on the cellular computation paradigm. These languages may provide a powerful tool for researchers and engineers that need to implement real-life applications on parallel machines using a fine-grain approach. This approach allows designers to concentrate on "how to model a problem" rather than on architectural details as occurs when people use low-level languages that have not been specifically designed to express fine-grained parallel cellular computations. In a sense, parallel cellular languages provide a high-level paradigm for finegrain computer modeling and simulation. While efforts in sequential computer languages design focused on how to express sequential data, objects and operations, here the focus is on finding out what parallel cellular objects and operations are the ones we should want to define. Parallel cellular programming emerged as a response to these needs. © 2010 Springer-Verlag Berlin Heidelberg.