Tutorial: Evolutionary Design Methods in Electronic Design Automation

accelerator, approximate circuit, electronic design automation, genetic programming, machine learning

The use of evolutionary algorithms (EAs) for the automated design of programs, electronic circuits, neural networks, and other computational structures has become a fruitful approach in the last two decades. The advantage of EAs is that they can handle the design process in a holistic, multiobjective way and create solutions with unique properties. This tutorial surveys the key ingredients of EAs and focuses mainly on genetic programming. It presents several techniques (such as incorporating formal verification methods and surrogate models) to improve the scalability of the method. Examples of evolved solutions (approximate arithmetic circuits, neural network architectures, and image filters) that show unique properties compared to conventional designs are presented and discussed.