Integrated Area-power Optimal State Assignment

This paper presents a state assignment algorithm with the objective of lower energyalong with area comparable to the area-targeting state assignments such as JEDI. Theunderlying framework is MUSTANG's complete weighted graph with weights representingstate affinity. The weight computation phase estimates the computation energyof potential common cubes using steady state probabilities for transitions. The weightcomputation phase also identifies a large set of potential state cliques, which are incorporatedinto a recursive bipartitioning based state assignment procedure. Reuse ofcliques identified by the weight computation phase results in a faster and efficient stateassignment. The energy targeting weights result in ≈9% lower area and 18% lowerpower than area targeting weights in JEDI over 29 MCNC Logic Synthesis 93 benchmarks.The clique based state assignment performs almost as well as the annealingbased state assignment in JEDI, and takes only about half as much time