Broadcom, Meta, Applied Materials, GlobalFoundries, and Synopsys are funding a US$125 million Semiconductor Hub at UCLA Samueli School of Engineering in California over five years. The program will support AI chip research and workforce training, with faculty and students collaborating with the companies on chip design software and manufacturing projects. Doctoral students will participate in yearlong internships with the partners, and the hub is designed to help move research into the market faster as chip complexity rises and the industry navigates uncertainty over how AI will reshape semiconductor development.
Program Structure
Faculty and students at UCLA will work directly with the five partner companies on chip design software and manufacturing projects. The program includes yearlong internships for doctoral students at partner organizations including GlobalFoundries, a contract chip manufacturer, and Applied Materials, a semiconductor equipment supplier. These internships give students experience with production challenges alongside academic research.
Broader Context: U.S. Chip Manufacturing Resurgence
The UCLA hub sits within a larger campaign to rebuild U.S. semiconductor manufacturing capacity. According to UCLA, U.S. semiconductor capacity dropped from 37% of global capacity in 1990 to 12% currently.
The initiative aligns with the National Semiconductor Technology Center (NSTC), a CHIPS Act research-and-development program. The NSTC's planned Design and Collaboration Facility is expected to be based in Sunnyvale, California, with more than US$1 billion in research funding planned for the state.
The region also includes the California DREAMS hub, a nearly US$27 million initiative led by USC with UCLA as a partner. This hub focuses on defense-related microelectronics for 5G and 6G, the next generations of wireless networks.
These hubs align with the CHIPS and Science Act's research-and-development goals, including the planned NSTC as a public-private consortium.
Why the Hub Model Addresses Industry Challenges
Chip technology continues to grow more complex, and the cost and risk of turning new ideas into products often exceeds what a single company can manage independently. The hub targets what researchers call the "valley of death"—the stage when laboratory prototypes fail because they were not designed for manufacturability.
The shared hub environment allows researchers to test risky ideas, including alternatives to current graphics processing unit (GPU)-based AI hardware. Many chip companies avoid funding such alternatives independently due to high failure rates. By distributing risk across multiple partners and leveraging academic resources, the hub model creates space for exploration that individual companies may not pursue alone.
