Behavioral level guidance using property-based design characterization

The growing importance of optimization, short time to market windows, and exponentially growing design complexity are just a few of the factors shaping the state-of-the-art synthesis process. In particular, optimization at the early stages of design is crucial--at the system and behavioral levels, orders of magnitude performance improvement in key design metrics such as throughput, power, and area can be attained. This requires, however, strategic and coordinated application of design techniques best suited for a target design. The problem, however, is the number of options currently available is overwhelming, and as a result, design exploration is often conducted in a qualitative, ad-hoc manner. To address these challenges, this thesis introduces a new design methodology for guiding the exploration process to quickly find effective sequences of design optimizations. The building blocks of the methodology are quantitative design characterization and a library of characterized optimization techniques. Design characterization is done using a set of techniques to automatically extract the "essence" of a design description. The library of characterized optimization techniques encapsulates knowledge about the effectiveness, scope, and interdependencies of various optimizations. These two building blocks enable analysis of optimization alternatives, and have been encapsulated in an interactive guidance environment. The guidance environment suggests and ranks potential optimizations, both in terms of immediate and longer-term impact. It also provides evaluations of the design and of the likely effects each optimization will have on performance. Using the provided guidance, designers can make decisions in a more informed manner and can explore the space more effectively, thus resulting in shorter design time and more highly optimized designs. A core contribution of this thesis is the design characterization. The essence of the design is captured using property metrics that are shown to be related to the quality of algorithm-architecture mappings. The following properties and their quantifications are presented: size, topology, timing, concurrency, uniformity, locality, and regularity. As well as being a key component of the guidance methodology, this work demonstrates the effectiveness of using property metrics in algorithm selection, performance estimation, and architectural synthesis.