Forecasting hardware resource requirements

This paper provides performance analysis and evaluation of a computer center in a large industrial organization; it is an extension of the Ahituv and Igbaria model. A regression model is utilized to forecast future hardware resource requirements given partial information. Several hardware resources were considered. Given the future growth in some of the resources and the number of jobs, the model can predict the growth in the other resources as well. Furthermore, the model can detect the resource that will be the bottleneck and at what level. The forecasts were performed for a large range of scenarios.