Advancements in Secure Computing: Exploring Automated Repair Debugging and Verification Techniques for Hardware Design

This paper studies the recent advancement of techniques for automating the repair and verification of hardware designs. Challenges associated with debugging and fixing bugs in hardware designs are discussed, emphasizing the importance of addressing these issues prior to the manufacturing of chips. It categorizes approaches according to their contributions and indicates specific techniques used in each approach. The studies demonstrate the effectiveness of different methodologies, such as fault localization, confidence scoring, Large Language Models, and learning-based approaches, in automatically repairing as well as hardware design verification. According to the results of each approach, these techniques have the potential to enhance the security and reliability of hardware systems.