Since its inception as the Integrated Circuits Laboratory more than a half-century ago, the Stanford Nanofabrication Facility (SNF) has been the go-to resource for Stanford researchers and industry collaborators across Silicon Valley working on the next generation of semiconductor-based computer chips. The Stanford Nanofabrication Facility – or SNF, for short – operates out of the Paul G. Allen Building, named for the late Microsoft co-founder. Today, a major renovation promises to keep the facility at the forefront of chip manufacturing through a new collaboration with and funding from the Taiwan Semiconductor Manufacturing Corporation (TSMC).
The announcement comes on the wave of major developments in the chip industry. The August 2022 rollout of the CHIPS and Science Act capped the Biden administration’s $150 billion initiative to bolster the American semiconductor industry. In June 2024, American chip designer Nvidia (founded by Stanford Engineering alumnus Jensen Huang, MS ’92) became the world’s most valuable company, a measure of global demand. Alongside these developments, a team led by Stanford engineers were awarded a multimillion-dollar Department of Defense contract to enable the translation of next-generation chips from university laboratories to fabrication (lab-to-fab) at large scale in industry.
“Suddenly, hardware is king … again,” says H.-S. Philip Wong, the Willard R. and Inez Kerr Bell Professor and professor of electrical engineering at Stanford, one of the world’s foremost chip technologists.
Wong is understandably optimistic about Stanford Engineering’s new collaboration with TSMC, which will place new research equipment in the Stanford Nanofabrication Facility. To support these new capabilities, the School of Engineering will co-invest in a 20% expansion of SNF’s lab space. This TSMC partnership caps a multiyear investment by the School of Engineering that refreshed aging equipment and updated facilities for improved energy savings.
“Stanford is in the heart of Silicon Valley,” Wong adds, “and we are the most visible in terms of research and advanced research in this space. Some of this equipment is valuable to our industry partners, as they facilitate new materials and semiconductor device research which is complementary to internal research at companies. The upgraded SNF is where it will happen.”
“It’s the beginning of a new era for the SNF, but it’s important to note that the SNF is not just a focus of semiconductor research for Stanford alone,” says Mary Tang, managing director of the SNF. “It serves more than 600 users annually, about 25% from outside Stanford in industry, government, and other academic institutions. All these organizations will rely on our state-of-the-art facilities and expertise to explore new possibilities in chipmaking for a variety of applications from bio-electronics to quantum computing.”
Opportunity ahead
Semiconductor technology is foundational to the way the world works now. Without its continued advancement, the promises of artificial intelligence, 5G communications, and quantum computing will never be realized. To that end, 12 of the 17 United Nations Sustainable Development goals depend on the continued advancement of information and communication technologies. All are based on semiconductors. Every day, almost two-thirds of the world’s population uses the internet.
Beyond sheer economic impact and national security, semiconductor technology plays a major role in research and development solutions to many of society’s greatest challenges – the digital transformation of health care, climate change mitigation, and protecting the environment among them. These solutions will all rely on computing and the massive deployment of the Internet of Things, with extreme energy efficiency and diverse functionalities. The demand for semiconductors is virtually endless. The only question is whether the research community can keep up.
Over the past five decades, semiconductor technology has enjoyed a seemingly never-ending path to advancement, providing faster, more energy-efficient computing and communication technologies at a rapid and regular pace. These advances have become so predictable that most of the world outside the semiconductor community may have grown complacent in the certainty that newer and better semiconductor technologies are always on the horizon, arriving like clockwork. But the past is never a guarantee of the future.
Hotbed of research
Semiconductor technology is now at an inflection point where it is poised to make the next leap by harnessing new materials, new devices, and new designs to significantly improve the energy efficiency of computing. The renovation will make the SNF one of the most advanced chipmaking research facilities in the academic world. SNF will be a hotbed of research on new materials and new devices.
In addition to manufacturing chips, TSMC – founded in 1987 by Stanford Engineering alumnus Morris Chang, PhD ’64 – is interested in exploring new materials and new devices, seeing great benefit in creating a shared facility where researchers in academia and industry work closely together to develop state-of-the-art technologies that are driving the field forward, explains Mark Horowitz, the Fortinet Founders Chair of the Department of Electrical Engineering and the Yahoo! Founders Professor in the School of Engineering.
“We find that new materials are essential for advancing semiconductor technologies. So the new equipment will explore new avenues for next-generation technologies. Industry is engaged in this effort because university facilities have the flexibility to handle new materials not currently used in manufacturing,” Horowitz says. “What I’m really excited about is that these upgrades will put us in a great position for the future with a state-of-the-art research facility that everyone can share.”
“Beautiful collaboration”
Srabanti Chowdhury, associate professor of electrical engineering, is among the SNF’s most frequent users. She and her lab are studying wide-bandgap (WBG) semiconductors beyond silicon, namely gallium nitride, gallium oxide, and diamond. These materials can offer game-changing advantages to our society where energy efficiency is always at a premium. For her, the SNF’s pioneering approach made her work – and that of many others – possible.
“The SNF set the standard of what a shared facility could and should be and many of us owe our careers to the work we were able to do here,” Chowdhury says. “There were many demonstrable breakthroughs in the lab-to-fab world that happened at SNF.”
The SNF upgrade will allow Chowdhury and fellow academic researchers and, importantly, students, to use cutting-edge equipment to which they would not otherwise have access. It is exciting to have hardware making headlines, she says.
Chowdhury's wide-bandgap materials and devices could transform electronics for power and radio-frequency applications. For instance, her group is pioneering new materials to enhance thermal management in electronics. The SNF upgrade will open new possibilities for her group to demonstrate the possibilities and the impact of such technologies.
“This is a very opportune time to be in chipmaking, whether it’s in academia or industry,” Chowdhury says. “The participation of industry is extremely important. In many ways, a shared facility is the only way the cleanroom works. We benefit from industry expertise, and they benefit from us demonstrating the technology, a proof-of-concept. It’s a beautiful collaboration.”
Less is Moore
Eric Pop, the Pease-Ye Professor and professor of electrical engineering at Stanford, is a semiconductor expert. His lab, including a team of postdoctoral scholars and doctoral candidates, is exploring semiconductors just a few atoms in thickness. They have seen hugely increased interest from industry collaborators in these materials and devices in the past few years, as companies are seriously looking for what’s beyond the ubiquitous silicon transistors in use today.
“If a 12-inch silicon wafer were the diameter of the continental U.S., the smallest features would be around 3 inches long – about the length of my index finger,” Pop says, making real the scales in modern semiconductors today.
Pop is eager to take advantage of SNF’s brand-new machinery. For Pop and his group, this is as much a matter of being able to reliably repeat a well-crafted experiment as it is about new, cutting-edge capabilities. The cleanroom is called a cleanroom for a reason – chip components are so small a single speck of dust can doom a promising experiment.
Such upgrades will allow Pop’s team to continue their pursuit of the boundaries of Moore’s Law, which predicts that computer hardware will double in density every few years or so. The fundamental limit of Moore’s Law has been anticipated since the 1980s, Pop says, but so far it has not yet been reached.
Pop is pursuing two-dimensional semiconductors and phase-change materials for energy-efficient computing and data storage. In their most innovative work, his team is working on integrating such materials into next-generation three-dimensional chips, where components are stacked atop one another like apartments in skyscrapers.
“Making 3D chips is extremely hard, but if we succeed – when we succeed – it will drastically shorten vertical connections between logic and memory devices,” Pop says. “We are looking forward to taking advantage of this new equipment and the essential reliability and cleanliness that SNF’s new machines and facilities will offer.”