Georgia Tech Addressing the Nation’s Call for Semiconductors

Semiconductors, or microchips, are vital to life in the modern world. They’re used in the microwave you heated your breakfast in this morning, the car you drove to work, the mobile phone you shouldn’t use while driving, the bank ATM you visited, and the screened device you’re reading this story on.

They’re in our TVs, refrigerators, and washing machines, helping us live comfortable lives. They also help us stay alive as part of the medical network, used in pacemakers, blood pressure monitors, and MRI machines, among other things. Also, our national economic and defense systems rely on them. Basically, semiconductors control and manage the flow of information in the machinery that keeps the world going.

And right now, at Georgia Tech, researchers are working to innovate chip technology to ensure that U.S. semiconductor development is globally competitive, reliable, sustainable, and resilient, today and in the future.

“If you look at semiconductors, or the whole area of computing, it spans across Georgia Tech — across many different schools and disciplines,” said Arijit Raychudhury, professor and Steve W. Chaddick Chair in the School of Electrical and Computer Engineering (ECE). “Starting with physics and chemistry, where we essentially learn how different types of materials will react, to materials science and engineering, to electrical engineering and computer engineering, to computer science.”

It's a diverse, multidisciplinary enterprise from bottom to top, Raychudhury noted. And there is still plenty of room at the bottom, as theoretical physicist Richard P. Feynman famously said more than 60 years ago, predicting that one day we’d be making things at the atomic level. We are. It’s a familiar realm to Victor Fung and his lab, where they are designing new materials for semiconductors from the ground up, atom by atom.

“We are interested in exploring how to translate the latest advances in AI and machine learning to aid in accelerating computational materials simulations and materials discovery,” said Fung, assistant professor in the School of Computational Science and Engineering. “We’ve been developing methods which can accurately predict a wide range of materials’ properties, to greatly facilitate high-throughput materials screening.”

Fung’s lab is using AI to discover previously unstudied materials with the electronic properties to build into chips. This approach to creating “designer” semiconductors would be significantly faster and cover more of the materials space than current methods.

Improving the Landscape

Smaller, more efficient, and more powerful are all part of the constantly evolving landscape in semiconductor research and development. It’s a very expensive landscape. While many chips are about the size of a fingernail, they are among the most complex human-made objects on Earth. Just building a semiconductor fabrication factory costs billions of dollars.

For a chemical engineer like Michael Filler, that sounds like opportunity.

“Chemical engineers think about how we produce products on a massive scale,” said Filler, associate professor in the School of Chemical and Biomolecular Engineering and associate director of the Institute for Electronics and Nanotechnology (IEN).

Filler, whose research involves the growing of semiconductor components, like transistors, from seed particles, is aiming to help democratize the process of chip development, bringing down the cost substantially while maintaining performance. In a not too distant future, that could mean an individual at home printing a chip on a machine similar to a 3D printer.

“Imagine a laser printer that can literally spit out custom electronics in a matter of minutes,” Filler said. “We’re big believers in the individual’s ability to be creative and know what they want to build for their applications. Ultimately, we’re interested in giving makers and prototypers opportunities to customize electronics.”

He’s in the right place for the far-reaching research he has in mind, adding, “We are so blessed with great facilities at Georgia Tech. It would be hard to imagine working somewhere else, because very few places have the diversity and quality of tooling we have here.”

IEN, which facilitates much of the semiconductor research at Georgia Tech, is based in the Marcus Nanotechnology Building, with its state-of-the-art micro/nano fabrication facilities such as the shared cleanroom space and a laser machine lab for micromachining.

But it is the range of expertise and creativity among faculty and students who are making IEN and Georgia Tech a thought leader in semiconductor research. This is evidenced by Tech’s recent grant of $65.7 million from the Semiconductor Research Corporation and the Defense Research Projects Agency to launch two new interdisciplinary research centers.

Events like Georgia Tech Chip Day (May 2) and Nanowire Week, an international gathering happening in Atlanta in October, also speak to Tech’s growing influence in this area.

Answering the Call

The Covid-19 pandemic clarified just how difficult it can be to make more chips. A shortage of semiconductors affected the supply of phones, computers, and other commonly used items during the global shutdown. Increased demand, depleted reserves, and too few manufacturing plants and workers significantly crippled the supply chain.

“The high degree of geographic concentration in certain parts of the semiconductor supply chain has recently created a heightened risk of supply interruptions,” said Chip White, Schneider National Chair in Transportation and Logistics and professor in the H. Milton Stewart School of Industrial and Systems Engineering (ISyE). “Such interruptions and resulting wild fluctuations in semiconductor demand can threaten the nation’s public health, defense, and economic security.”

With that in mind, translational supply chain research is going on in several places on campus, White said, including the Supply Chain and Logistics Institute and the NSF AI Research Institute for Advances in Optimization. White and his colleagues are developing software platforms for stress testing manufacturing supply chains. The goal is to identify vulnerabilities and risk mitigation procedures to design and operate next generation supply chains for critical industries such as the semiconductor industry, to improve global competitiveness and strike a balance between market forces and national security.

In an effort to address and feed the next generation demand for chips, the Biden administration recently launched a massive effort to outcompete China in semiconductor manufacturing, offering $39 billion in funding incentives for companies seeking to build plants in the U.S.

Another related area of importance in the ongoing development of semiconductors is growing the workforce of the future, and that includes a new wave of researchers. This is a role that Jennifer Hasler takes seriously.

“I have a strong interest and belief in mentoring,” said Hasler, ECE professor and founder of the Integrated Computational Electronics lab at Georgia Tech. She’s proven, theoretically at least, that the technology already exists to build a silicon-based version of the human cerebral cortex (which would cost billions of dollars to design and build), but one of her favorite roles is working with new, young faculty.

“It’s a personal thing for me, but it’s one of the coolest things I’m involved in,” she said. “When they come to Georgia Tech, they see how big this place is, bigger than a company. I like to say to them, ‘Let’s calm down, take a breath, you’re good, so let’s go make some cool stuff. Let’s get some momentum going.’”

For Raychowdhury, director of the new Center for the Co-Design of Cognitive Systems (part of the JUMP 2.0 program), developing the skilled workforce of the future means answering the call of the nation.

“This is one of the largest ECE departments in the country, with many, many talented students,” he said. “And given the need and shortage of skilled professionals in this particular area, I think it’s critical for us to create that kind of pipeline.” Last year, ECE undergraduate students started taking a new, two-semester course, sponsored by Apple, in which they actually build microprocessors from scratch.

“This is completely new,” Raychowdhury said. “It’s expensive to offer this course, but we plan to keep doing it and we’re in conversations with other companies that want to invest in workforce development. So, in addition to doing fantastic research, we want to be sensitive to the needs of the country and a new generation.”

Writer: Jerry Grillo