CICI: UCSS: Trustworthy Cyberinfrastructure-Enabled Data-Driven Electronic Design Automation for Collaborative Design of Analog/Mixed-Signal Integrated Circuits

Electronic design automation (EDA) underpins the development of microchips that power various technologies from healthcare and communications to transportation and national security. As chips grow more complex, especially analog and mixed-signal (AMS) systems, current design methods are increasingly slow, costly, and difficult to scale. Artificial intelligence (AI) offers a promising path to accelerate chip design, but relies on highly sensitive proprietary data, raising major concerns about privacy, security, and trust, which limit collaboration and slow innovation. This project addresses these challenges by enabling secure and trustworthy AI-driven chip design, allowing organizations to collaborate without exposing confidential data. By improving efficiency and scalability, the project will strengthen U.S. semiconductor leadership, support economic growth, and advance national defense. It will also provide open tools and training to build an interdisciplinary workforce in hardware, software, and AI. This project introduces TrustEDA (Trustworthy Data-Driven EDA), a paradigm that integrates data-driven EDA with cyberinfrastructure (CI) while embedding privacy, security, and robustness as core constraints. Focusing on AMS design, it develops scalable, collaborative CI-native workflows through three thrusts: (I) CI-ready datasets, models, and privacy-preserving deployment; (II) domain-specific security analysis and defenses against data leakage and model exploitation; and (III) CI integration and generalization to enable reusable, scalable CI workloads for large-scale design-space exploration and continuous model improvement. The project also delivers OpenAMS, an open-source platform with datasets, tools, and benchmarks to promote reproducible research and accelerate innovation in AMS circuit design, EDA, and AI, establishing a trustworthy foundation for next-generation data-driven chip design. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2613629 | Program: 01002627DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Weidong Cao | Institution: George Washington University, WASHINGTON, DC | Award Amount: $600,000 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2613629 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2613629.html