Investigating grid computing technologies for use with commercial simulation packages

Abstract

As simulation experimentation in industry become more computationally demanding, grid computing can be seen as a promising technology that has the potential to bind together the computational resources needed to quickly execute such simulations. To investigate how this might be possible, this paper reviews the grid technologies that can be used together with commercial-off-the-shelf simulation packages (CSPs) used in industry. The paper identifies two specific forms of grid computing (Public Resource Computing and Enterprise-wide Desktop Grid Computing) and the middleware associated with them (BOINC and Condor) as being suitable for grid-enabling existing CSPs. It further proposes three different CSP-grid integration approaches and identifies one of them to be the most appropriate. It is hoped that this research will encourage simulation practitioners to consider grid computing as a technologically viable means of executing CSP-based experiments faster.