Analytics in Design Verification and Test - Principles, Opportunities, and Challenges

Electronic Design Automation (EDA) and test have become major areas for applying data mining in recent years. In design and test processes, tremendous amounts of simulation and measurement data are produced and collected. These data present opportunities for applying data mining.
In this talk, I will first explain the working principles of data mining in design verification and test. Application examples will be presented to explain how a problem can be formulated to facilitate the application of these learning techniques. Opportunities for applying data mining include problems in areas such as functional verification, layout analysis, timing analysis, design-silicon correlation, test cost reduction, diagnosis, and customer return analysis. In these applications, I will take the view that data mining is to enhance one's domain knowledge using data. I will highlight the challenges in this view and provide a summary of the recent application results in practice. Experience of developing a data mining flow will be presented and promises of applying data mining will be demonstrated, both through results in selective applications based on industrial settings.
Keywords: Computer-Aided Design, Data Mining, Machine Learning, Test, Verification, Validation