Drexel University’s College of Engineering is part of a $10 million research and development initiative to advance the use of artificial intelligence in microchip design for next-generation wireless communication.
The project, funded by the National Semiconductor Technology Center (NSTC) and led by Princeton University, is one of several major national initiatives established under the CHIPS and Science Act to strengthen U.S. leadership in semiconductor research and innovation. Over the next 30 months, the Princeton-led team—including academic and industry partners such as Drexel, the University of Southern California, Northeastern University, Cadence, Raytheon, Keysight, and GlobalFoundries—will work to automate the traditionally labor-intensive design of analog and radio-frequency (RF) integrated circuits. These components are critical to devices ranging from satellites to self-driving cars.
Ioannis Savidis, PhD, associate professor of electrical and computer engineering, is leading Drexel’s contribution to the effort. His research group is developing machine learning models that can predict the performance of circuit designs based on their structural parameters, reducing reliance on costly and time-consuming simulations.
“We’re building forward prediction models that analyze circuit data, material properties, and design flows to estimate performance metrics of RF circuits — things like gain, bandwidth, power efficiency, match network structure, and more,” Savidis said. “That’s a critical piece of building scalable AI tools that can assist in the chip design process.”
In addition to ML-driven prediction, Drexel’s team is advancing models that can generate novel RF circuit topologies. These topologies are automatically sized and biased by machine learning algorithms—a task that currently requires the expertise of highly trained RF design engineers.
Drexel’s share of the funding for this project totals just over $1 million. Work is already underway, with Drexel researchers supporting the larger effort to demonstrate how AI can lower development costs, accelerate design timelines, and unlock new architectures that surpass traditional performance benchmarks.
Beyond its role in the NSTC, Drexel is also a member of the Air Force Research Laboratory (AFRL) Mid-Atlantic Hub Network, a Department of Defense initiative focused on transitioning microelectronics innovations from research labs into manufacturing and real-world applications. Membership in both NSTC and AFRL positions Drexel to compete for future funding to prototype and commercialize semiconductor technologies alongside other regional academic and industry partners.
“These national networks are giving us access to emerging opportunities across the semiconductor ecosystem,” Savidis said. “Whether it’s through NSTC or AFRL, we’re putting ourselves in position to contribute to the next generation of chip design, with real pathways to commercialization.”
Together, these affiliations highlight Drexel Engineering’s expanding role in the national effort to reestablish U.S. leadership in semiconductor research, design, and workforce development. As the Princeton-led project progresses and new opportunities emerge through AFRL, Drexel’s faculty and students are poised to help shape the technologies powering tomorrow’s RF communication infrastructure.