Abstract
This chapter deals with the usage of grid technologies for nature-inspired algorithms and complex simulations. After shortly introducing the grid and its technological state of the art, some features are pointed out in order to set the boundaries of the applicability of such new technology to the matters of interest. Then two paragraphs show some possible usages of grid technologies. The first one introduces the master-worker paradigm as a conceptual and technological scheme that helps in solving issues related to dynamic optimisation via nature-inspired algorithms and in exploring the parameters space of complex simulations. The following paragraph concerns two other points: the possibility to distribute agents of agent-based simulations using multi-agent systems; and the boundaries, architectures, and advantages in distributing parts of complex simulations which are heavy from the computational point of view. The chapter, as a whole, acts as a guide presenting applicative ideas and tools to exploit grid technological solutions for the considered purposes.
Key Terms in this Chapter
Message Passing Interface: A standard interface that allows access to memory areas of distant computers. Generally implemented by low-level programming languages, it also permits the communication between processes in execution on different computers.
Open Grid Services Architecture: A set of principles defined by standards organisations for developing Grid middleware capable of allowing grid systems based on Web services.
Master-Worker Paradigm: A computing architecture operating on a network. A single computer is called the master and sends jobs to be executed by other computers, called workers. The communication between master and workers generally happens just at the beginning and at the end of each job.
Grid: A new computing infrastructure that takes inspiration from power grids and that aims to make available computing services, such as computer power and data storage, as a commodity accessible from anywhere.