Design flow and methodology for 50M gate ASIC

This paper presents a methodology for full chip RTL timing closure for very large ASICs. The methodology is based on the concept of a "silicon virtual prototype". The methodology is based on the scalable technique of clustering and cluster placement and leverages the tight integration between the algorithms by means of a common, unified data model.