Forecasting Microprocessor technology in the multicore era using TFDEA

Technological advancements in the microprocessor industry are benchmarked and gauged against a set of diverse criteria, specific to the fabrication process, usage as well as achieved performance. Changing trends in the appeal factor as well as wide variety of growing application of microprocessors in different industries also have a defining impact in the advancement of the technological features in future. This study improves the previous investigation in forecasting microprocessors' technology and uses Technology Forecasting using Data Envelopment Analysis (TFDEA) methodology for an enhanced model. The study takes advantage of the recent microprocessor dataset including multi core processors, from a variety of resources including the dataset collected by Stanford University, the database made available by Standard Performance Evaluation Corporation (SPEC), and the specifications announcements by the microprocessor manufacturers such as Intel and AMD. The result of this study is a rate of change (RoC) that is obtained based on the recent design trends including the State-of-the-Art generation of multi-core microprocessors and hence, is superior for forecasting the future microprocessor technology trends. The Rate of Change obtained provides the rate in which values of expected output performance or input requirements for the state of the art microprocessors change in future years and can be used to evaluate the competitiveness of the projects being researched and developed.