Grid Economics in Departmentalized Enterprises

The application of Grid technology is finally spreading from engineering and natural science-related industrial sectors to other industries with a high demand for computing applications. However, the diffusion of Grid technology within these sectors is often hindered by a lack of the incentive to share the computational resources across departments or branches even within the same enterprise. A promising way of overcoming these barriers is the introduction of a pricing mechanism for the use of Grid-based resources. This work introduces such a pricing approach to Grid computing and provides three simulation scenarios to illustrate the effectiveness of such an economized Grid solution. The simulation results indicate that the pooling of information technology resources can produce a reduction of 33% in cost compared to individual and dedicated servers. However, with a price-based allocation of computing resources, a further 10% of cost reduction can be achieved by introducing an auction mechanism. Therefore we claim that there is huge cost reduction potential in departmentalized enterprises beyond the savings that can be achieved by a utility-based allocation of computing resources, if economically measured allocation methods are combined with advanced refining and learning methods in the allocation process.