Elements: AnalogEDA-Hub: At-Scale Analog Circuit Electronic Design Automation with An Open and Efficient Platform

Foundational semiconductor integrated circuits (ICs) research provides a core national competitiveness by driving both trillion-dollar economic growth and advances of various transformative technologies. Thus, advancing Electronic Design Automation (EDA) tools to enable agile, low-cost, and high-quality development of ICs is a national research priority. While digital ICs have enjoyed mature automated design flows to cope with the need for higher productivity and stronger capabilities, analog ICs have historically relied on manual design processes despite their essential role in ubiquitous electronic systems. This unsustainable, handcrafted process has been a major bottleneck to cost reduction and productivity improvements in the current IC ecosystem. While analog EDA has long been sought after by academia and industry since the 1950s, there are still no mature methods widely embraced. This is largely due to the lack of foundational cyberinfrastructure to address the scalability of the current analog EDA research paradigm. The need for such a cyberinfrastructure has become increasingly urgent due to the recent shift toward AI/ML-enabled analog EDA research, which is inherently data-driven and benchmark-focused. This urgency is further amplified by the unprecedented demand for and scale of analog ICs, driven by the growing complexities of emerging technologies such as generative AI, 5G/6G communication, and quantum computing. This project will develop AnalogEDA-Hub, the first-of-its-kind open and efficient platform for AI-driven analog EDA research at scale. AnalogEDA-Hub serves as a comprehensive open-source infrastructure, offering multi-modal datasets, learning-based tools, and standardized design environments for versatile and representative analog circuits. It aims to facilitate scalable, generalized, and holistic development of diverse analog EDA methodologies across communities. The platform enables users to efficiently develop and benchmark various EDA methods in a structured, interactive, and reproducible manner, while also accelerating the generation/discovery of novel analog circuits to sustain the performance of analog ICs in the post-Moore era. Fully compatible with existing learning frameworks and conventional optimization algorithms, AnalogEDA-Hub ensures broad accessibility and usability, significantly expanding its impact across diverse research and development communities. In addition to advancing research, this project integrates a robust outreach plan and fosters collaboration among stakeholders to train the next generation of talent in the IC, EDA, and AI/ML fields. By engaging graduate, undergraduate, and high school students in education and training, AnalogEDA-Hub will play a vital role in building a skilled workforce for future technology innovation and democratizing domestic analog IC development. 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: 2514071 | Program: 01002627DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Weidong Cao | Institution: George Washington University, WASHINGTON, DC | Award Amount: $599,971 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2514071 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2514071.html