Abstract: A method and apparatus for ensuring efficient validation coverage of an architecture, such as protocol or interconnect architecture, is herein described. A coverage space of states for an architecture is generated and stored in a database. During simulation, states of the coverage space encountered are marked. From this, the states encountered and not encountered may be determined. Based on the states not encountered, a targeted test suite is developed to target at least some of the states not encountered during previous simulation. This feedback loop from simulation to refining of a test suite based on states of a coverage space not encountered during simulation may be recursively repeated until adequate validation, i.e. an adequate confidence level of validation, of the coverage space is achieved.