We are currently examining the use of Finite State Machines to model processor pipelines. Our aim is to discover the extent to which the complexities of modern superscalar pipelines are amenable to theoretical modelling. As a result, we hope to improve our understanding of pipeline design. In this short paper we illustrate our approach by analysing the trade-offs between separate and unified caches in a superscalar system.
|UH Computer Science Technical Report
|University of Hertfordshire