Final Report on Practical Data Learning Tools and Methodologies for Verification
Research Report Highlight
This final report from UCSB describes data mining approaches to improve the efficiency and quality of functional verification for processors, reducing simulation cost by 80-95% and improving verification coverage.
This final report summarizes the major accomplishments from our project. We also point out three directions of future research. The first is to extend the approaches to go beyond simulation, such as emulation. The second is to extend the approaches for scenario generation. The third is for security verification. We will explain how these diverse applications can be addressed with the data analytics approaches developed in our research.