ABSTRACT
This conclusion presents some closing thoughts on the concepts covered in the preceding chapters of this book. The book focuses on the state-of-the-art of variation-aware design and the technologies which have become more and more complex and the available techniques for circuit designs as well. The problems in modern custom IC design are not due to a single law, single parameter, or single technique. The book shows that statistics can be treated by many different techniques and not only by Monte-Carlo. Disproving that the design is correct is quite an easy task for automated tools, but a positive proof is more difficult. One way to improve dealing with complexity is to exploit hierarchy; in optimizations, designers often do so. Besides all improvements, pure optimization is not efficient for all kind of problems. It may work out well for migration between foundries having a similar process, but for bigger changes in technology or specs, optimization should be augmented with migration scripts.
