UCLA’s $125M Semiconductor Hub: “We Want High Impact, Not Incremental Research”

UCLA Samueli School of Engineering launched a $125 million semiconductor hub with Broadcom, Applied Materials, GlobalFoundries, Meta, and Synopsys to accelerate AI-powered chip research and workforce development. Dean Alissa Park emphasized high-risk, high-return research over incremental gains, while Synopsys' Shankar Krishnamoorthy called the collaboration essential to overcome compute, memory, and interconnect walls by 2030.

Last month, the UCLA Samueli School of Engineering announced it had partnered with Broadcom, Applied Materials, GlobalFoundries, Meta, and Synopsys to establish a $125 million semiconductor hub aimed at accelerating research and workforce development in AI-powered chip technologies. The headline investment figure is an initial five-year commitment of philanthropic gifts $25 million from each company and in-kind support i.e., technology to establish a long-term collaboration across the semiconductor ecosystem, spanning chip design, software, manufacturing, equipment, and advanced materials. The mission of the UCLA semiconductor hub is to foster sustained, mid- to long-term collaboration between leading faculty and industry partners to drive advances in connectivity, computing, and intelligent systems. In an exclusive interview with EE Times this week, Ah-Hyung “Alissa” Park, the Ronald and Valerie Sugar Dean of Engineering at UCLA Samueli, said this program was not just looking for incremental research impacts. “What I really want is high risk, high return impact .” She emphasized in the conversation that when students had a great idea, she did not want them to wait three years to get there—and that they would be encouraged to try those ideas immediately. And if the idea was deemed not quite right collectively by the students and the industry-academia teams involved, they would be allowed to pivot quickly to enable high-impact results. View All https://www.eetimes.com/category/sponsored-content/ Listen to the full audio recording of the conversation with Dean Alissa Park of the UCLA Samueli School of Engineering and Shankar Krishnamoorthy, chief product development officer at Synopsys: In the interview, we were joined by Shankar Krishnamoorthy, chief product development officer at Synopsys, who emphasized that the UCLA semiconductor hub is less a nice-to-have and more a necessity to avoid hitting a wall by 2030. He framed the current moment as an “unprecedented infrastructure build-out” for the AI token economy, where agentic workflows https://www.embedded.com/a-look-at-agentic-ai-in-the-eda-engineering-workflow/ run across every industry at acceptable cost and power. To get there, he argued, the industry needed to break through three hard limits at once: the compute wall ever higher teraflops for bigger models , the memory wall moving data fast enough between memory and compute , and the interconnect wall linking vast numbers of compute and memory elements efficiently . Left to organic, incremental innovation alone, he does not believe the sector can move fast enough, especially under escalating power and energy constraints. That is why he sees this kind of deep, structured collaboration with academia as essential rather than optional. For him, the UCLA hub works because it is full-stack and vertical—spanning materials, manufacturing, design tools, chips, and large-scale systems, with companies such as Applied Materials, GlobalFoundries, Synopsys, Broadcom, and Meta all at the table. Rewriting the rulebook on industry-academia collaboration When we spoke to Dean Alissa Park, you could sense both a huge passion for doing things differently and a sense of urgency. Gone are the days of siloed engineering departments and for academics to wait years to get their tenure through published papers. She argues that impact should be measured differently and that innovative research talent should be both nurtured and retained, while also being given the flexibility to pursue opportunities in industry when that offers the most promising path for their work and development. She is effectively trying to rewrite the rulebook on how universities work with industry. Trained as a chemical engineer and a climate and decarbonization researcher, she came to UCLA with a clear condition: The engineering school had to be deeply connected to the real-world technology ecosystem around it. At the semiconductor hub, that philosophy shows up in hard structural changes—from how faculty are evaluated to how students are trained. Instead of judging impact purely by papers, Park wants tenure and promotion to explicitly recognize when a researcher’s work changes how companies design or build chips. For her, fundamental, applied, and commercial research can no longer be separate stages; in the AI and semiconductor era, they have to happen in parallel and stay tightly linked. She said she is also reshaping the talent pipeline. Ph.D. students in the hub are admitted as a cross-disciplinary cohort, co-advised by both a faculty member and an industry mentor. They spend their first three years at UCLA and then an entire fourth year embedded in a partner company—a level of integration she says has “never been done before at the Ph.D. level.” A key to this is to break departmental boundaries within the school to enable exchanges of ideas in interdisciplinary teams. Park said, “I wanted to create a hub on one thematic area that every single engineering department comes together, so there’s a wide range of highly interdisciplinary activity , because you know, in industry, I’m sure Shankar is not only hiring computer scientists, there are all kinds of engineers, we just call them engineers, we don’t categorize by department, so if we can have all the engineers working together and bring other disciplines, of course, but if companies are kind of becoming a one team and we generate new knowledge and translate the knowledge into the market, but also educate students and train them in a meaningful way from day one, why not?” Aiming for high-risk, high-impact, and the ability to pivot quickly On the funding side, Park is explicitly pushing for high-risk, high-return projects, willing to back bold ideas even without preliminary data and to pivot quickly if a hypothesis does not pan out, while keeping students protected and fully supported. To make the money work harder, she persuaded UCLA to waive its usual 57.5% overhead on this hub so that almost every dollar from the founding companies flows into research, technical staff, and infrastructure. Coupled with a dual-leadership model—a faculty director and an executive director drawn from industry—Park’s approach aims to make sure the hub is a fast-moving joint venture between academia and the semiconductor supply chain. Defining direction: the Scientific Advisory Board The semiconductor hub has established a Scientific Advisory Board, a flexible body of technical leaders from each founding company plus invited external experts whose purpose is to define and continuously refine the hub’s research agenda, align UCLA projects with real industry roadmaps, and help drive high‑impact, high‑risk work from concept through validation into production. Synopsys’ Krishnamoorthy explained that this scientific advisory board he helps lead is designed to sharpen a small number of ambitious research thrusts, such as high-speed networking, co-packaged optics, and advanced heterogeneous integration, and then give those ideas the best possible shot at success through shared tools, prototypes, and rapid validation. He suggested that this model is distinctive in its “all‑in” posture: faculty, graduate students, and industry engineers iterating together, with a clear route for successful concepts to move from lab to production. He added that without that kind of risk-taking, multi-disciplinary, materials‑to‑workloads collaboration, the industry simply will not reach the compute and efficiency targets that AI build-out demands by the end of the decade. Coupling research, risk, and real-world deployment Speaking with Dean Park gives you the feeling that she is both ambitious and focused on ensuring UCLA delivers a model that meets industry needs, rather than just pursuing research on its own. By tightly coupling fundamental research, high‑risk experimentation, and real‑world deployment across the full stack from materials to workloads, the hub is aiming to address the compute, memory, and interconnect bottlenecks that everyone in the industry is talking about right now. The model of industry‑co‑designed research, an integrated Ph.D. pipeline, and a curated ecosystem of leading semiconductor players takes them from generating technical papers to translating breakthrough ideas into manufacturable technologies, and at the same time, developing a deeper talent pool for the semiconductor industry. See also: Clever Cooling Device Shows Imagination, Perhaps Even Utility https://www.eetimes.com/clever-cooling-device-shows-imagination-perhaps-even-utility/ As Chips Go Vertical, Metrology Struggles to Keep Up https://www.eetimes.com/as-chips-go-vertical-metrology-struggles-to-keep-up/