Verification data analytics with machine learning

Verification is data-and computation-intensive, making it an ideal field for ML applications. Advancements in ML have offered many opportunities to accelerate verification workflow, improve verification quality, and automate verification execution. However, being a data-centric method, ML has also elevated data to become the most crucial factor of ML success.

This whitepaper provides an overview on the importance of data to ML, the available data for verification, and the existing applications of ML in verification. It reveals that data itself may dictate applicable ML models. Machine learning has demonstrated great potential in verification. However, attention should be paid to generalizing and scaling the models to ensure their success in a production environment. And a data strategy to build the verification data assets will ensure the long-term success of applying ML in verification.

Introduction

Ever-increasing design complexity and shortening design-to-market time has demanded faster and more accurate functional verification. Industry surveys indicate that design engineers spend about half of their time on functional verification, and the situation has not improved over the years.

Increasing efforts have been spent on improving verification performance to reverse this trend. With more data gathered from an IC design’s life cycle, it is now possible to gain unprecedented insight by analyzing the data with machine learning (ML).

Recent advances in ML, especially the emergence of large ML models, afford the possibility of gaining knowledge in solving verification problems beyond individual projects or designs. Verification is a data problem, whereas ML is a powerful tool dramatically changing how verification can be done.