Chip Industry Talent Shortage Drives Academic Partnerships

Universities around the world are forming partnerships with semiconductor companies and governments to help fill open and future positions, to keep curricula current and relevant, and to update and expand skills for working engineers.

Talent shortages repeatedly have been cited as the number one challenge for the chip industry. Behind those concerns are several key drivers, and many more domain-specific requirements.

Among them:

• More electronics everywhere, and more intelligence in those electronics, requires new chip architectures, more hardware-software co-design, and more cross-training of engineers in disciplines that traditionally have been siloed;

• Geopolitical rivalries and pandemic-related supply chain glitches are prompting governments to pour massive resources into training and education in an effort to on-shore and re-shore key components and materials, and the expertise to design, verify, manufacture, and package advanced chips, and

• Cybersecurity threats in safety- and mission-critical applications, including longer lifetimes for chips in markets such as automotive and medical, require much deeper understanding of how to build secure devices and how to keep them updated and resilient. This is compounded by the growing number of connected devices, which significantly widens the attack surface.

The required bank of skills increasingly includes multi-physics design and verification; 3D integration; IC manufacturing process development, testing and inspection, and an understanding of how chips and chiplets can operate in an advanced package. In addition, there are more partnerships focused on developing technology related to AI/ML, 5G, quantum computing, and sustainability.

“Technology follows an exponential time scale,” said Keith Schaub, vice president of technology and strategy at Advantest America. “Engineers follow an evolutionary time scale, which is very slow. So even though we had hundreds of thousands of years’ head start, eventually it catches up to you. When I graduated, all you needed was a bachelor’s degree to get a job. The next generation needed a master’s degree. And then, after that, it’s a Ph.D. When I look around today, it’s like you need multiple PhDs, and you need to be able to add a new one every other year. That’s not sustainable. We keep adding these new challenges, like heterogeneous integration and chiplets, and while everything engineers learned in the last 10 years is useful, it’s not enough.”

Industry leaders, governments, and education recognize the need for change and work is underway to create the broadest partnerships for developing and updating skills in the history of electronics. (See Partnerships Table).

Workforce development
The shortage of skilled workers is an urgent issue for the chip industry and many partnerships either have been formed, or are being formed, to address this need. For example, Arizona State University (ASU) partnered with Advantest and NXP on a course about radio frequency test to train engineers for the chip test industry.

“Fundamentally, these conversations with industry start and end with workforce,” said Kyle Squires, ASU’s dean for the Ira A. Fulton Schools of Engineering and vice provost for engineering, computing, and technology. “And the need within the new fabs that are being built is monumental. If you really want to drive a lot of innovation, you need more individuals achieving engineering degrees and technician training opportunities. The more you have of them, the more ideas that you can drive to new innovations. And if we get that right, then we’ll have really made a difference.”

Rochester Institute of Technology (RIT) won one of six grants from the U.S. National Science Foundation (NSF) to provide undergraduates with hands-on research opportunities in STEM priority areas related to semiconductors, and received a donation of $500,000 over 10 years from onsemi to increase the pipeline of engineers in the computer chip field.

“These government/university/industry partnerships are critical,” said Doreen Edwards, dean of RIT’s Kate Gleason College of Engineering. “With the CHIPS Act, we are expecting many, many jobs in this area and it is going to take all of us ramping up our educational efforts to fill that workforce pipeline. It’s really all hands on deck. And it starts with the youngest students, getting them excited about science and getting them interested in and aware of the multiple career opportunities that science and engineering provides. Those partnerships will extend beyond just university, industry and government, but we’ll also start to include community partners, and the general public.”

Cornell University joined the Semiconductor Education Alliance aiming to address the growing challenges of finding talent and upskilling the existing workforce. Partners include universities from India and Europe along with Arm, Cadence, Semiconductor Research Corporation (SRC), and the Taiwan Semiconductor Research Institute. “We need more people, more students and workers, because this is an ecosystem,” said Huili (Grace) Xing, the William L. Quackenbush professor of engineering in materials science and engineering, and in electrical and computer engineering, and director of the JUMP 2.0 SUPREME Center. “You cannot expect a very few people to sustain the ecosystem, no matter where you’re talking about.”

Evolving curricula
So how can universities train students for a continuous and rapidly changing technology? This is especially difficult because it involves both software and hardware, and more domain-specific and increasingly heterogeneous architectures. And regardless of whether these devices are tethered to a battery or plugged into a socket, they need to be much more energy-efficient. Given the slowdown in Moore’s Law and the shrinking power, performance and area/cost benefits of scaling, that often requires a mix of computer science, electrical engineering, and in packages, an increasing amount of mechanical engineering.

The bottom line, according to Advantest’s Schaub, is that no amount of classroom-based education is ever going to be sufficient. Instead, he said that digital tools will be required in the future, particularly generative AI based on large language models. In effect, engineers need better search tools, rather than relying on expertise that used to be handed down from one engineer to the next, because that no longer can happen fast enough or in the context of increasingly complex chip and system design.

Others agree. “Ultimately, the way curriculum evolution happens is because you have talented faculty who are driving research in specific areas,” said ASU’s Squires. “We’ve been in an aggressive mode of hiring these faculty and, well, guess what they do? They’re not only doing research on those topics but it also informs their teaching. So that begins to infuse the curriculum sort of organically with these very contemporary ideas into what we have now.”

To keep up with technology, ASU offers new courses, but also bite-sized elements of a curriculum that gives students an opportunity to become familiar and up-to-date with very contemporary topics.

“This is so much fun, frankly,” said Squires. “Mechanical engineers, electrical engineers, those disciplinary trainings through those curriculums, they’re accredited and we have a very vigorous process that will continue. But these smaller, bite-sized chunks of curriculum will allow a student to broaden. So as a mechanical engineer, I may not necessarily have either capacity in my studies, or the depth of interest, to take an entire course on heterogeneous integration. But I might be very open to a smaller, bite-sized piece that’s looking at the thermal properties of packaging and new effects occurring because of things like heterogeneous integration. And that is going to be very important for us to be more nimble, to get these things done more quickly. And it’s a huge opportunity. We’re trying to figure out the best way to get that really rolling. But that would be also a way to impact local workforce through lifelong learning.”

Squires added that one of the key outcomes of a four-year engineering degree in any specialty is that you’re giving students the mindset and the tools. “When new technologies are emerging in their workspace, they’re going to be able to pick that up,” he said. “So it’s part of building that mindset in that culture. But it’s also about providing these new avenues for much more rapid uptake of knowledge and skills, and that’s perhaps the most pressing need we’re facing right now.”

One of the big challenges is teaching students how to stay current. “We’re asking ourselves, ‘Okay, what are the needs for the job market today?’” said Cornell’s Xing, noting the university focuses on teaching the fundamentals well, and then keeps up with technology through a three-tiered approach to learning. “First, this is the conventional problem, and this is how you apply it. Then, this is how you apply it to a more modern problem. And last, we challenge a student with future problems. ‘Can you come up with solutions, because we’re still researching those and we don’t know the solutions? I can teach you what we did in the past and the recent past, but then I can also inspire you.'”

At that point, students can join interdisciplinary research groups and collaborate with colleagues on areas of interest.

In the Netherlands, Eindhoven University of Technology (TU/e) held an inaugural summer school in partnership with several universities from Taiwan and industry partners, including TSMC and ASML. The primary goal was to share expertise to create the most relevant content for students, which can also help with future recruitment.

“What we need is electric engineers as they were educated in the ’70s and ’80s, who can build a chip,” said Martijn Heck, TU/e’s professor in the photonic integration group of the electrical engineering department. Many students instead study software and now AI, moving further and further away from transistors and hardware. “But photonics is growing, momentum is growing with the EU Chips Act, and we attract people to the basics with CBL [challenge-based learning], or we co-pitch electronics and photonics to try to poach students from other specialties. Photonics is part of the semiconductor world and we see this technology maturing. We want to step away from physics, and we should also look to higher abstractions.”

Aida Todri-Sanial, professor in the integrated circuits group of the electrical engineering department and scientific director for the summer school, agreed with Heck. “Our specialty is CBL and we want students to think about problem solving from the bachelor’s level, then follow different tracks such as the AI track, semiconducting in various advanced nodes, automotive, and physics with quantum at the masters level.” While she said there is no official path to employment after study, industry is pushing the university to double its program, and graduates can be confident of being hired locally or abroad.

Curricula support from industry
To support academia amid the growing demand for workforce, many semiconductor companies also have long-standing university programs that are open to multiple institutions. Among them are Synopsys, Siemens EDA, Codasip, Ansys, Keysight, Arm, Teradyne, Imperas, and Cadence, whose programs offer access to software, tools, tech support, grants, curriculum, and online training. The industry also supports grade schools with a range of initiatives, such as Ansys’ recent partnership with Formula One to engage students from 58 countries in engineering competitions, introduce workforce skills, and inspire career opportunities.

One of the big problems with the current education curricula is that engineering for chips is becoming so complex and interconnected that just learning one facet, such as power or packaging or embedded software is no longer sufficient. This is particularly evident when it comes to embedded security, which entails everything from hardware and software to interfaces and I/O.

“You could hire somebody who has a background in electrical engineering or computer engineering, where they understand the low-level hardware and how to build embedded systems and how to develop them, but they don’t usually have a background in securing them,” said Dan Walters, principal embedded security engineer and lead for microelectronics solutions at MITRE. “Or you could look at students with more of a focus in security and cybersecurity. Those typically are computer science degrees. And some universities have computer or cybersecurity degrees, but that’s really software-heavy. Those students don’t understand embedded systems and the unique things that come along with that. What we essentially did was hire from one of those two groups and say, ‘Okay, we’re going to do on-the-job training for the other 50% that you’re missing.'”

Geopolitics and innovation
It’s been a complicated year for the global chip economy, with U.S. bans on exporting advanced AI chips followed by China’s ban on the export of key raw materials. Plus, there are still lingering concerns after the pandemic disrupted the supply chain.

“Those of us who have been in this field have recognized the risks of not having a domestic manufacturing base for microelectronics and semiconductors for some time,” said RIT’s Edwards. “All of the companies we worked with expressed this concern for years, and when the pandemic hit, we really saw how vulnerable we were. And this has really spurred a lot of government, university, and industry collaboration. It’s what we’ve been doing for 40-some years, but the volume of those relationships has increased substantially in the past couple of months.”

For example, RIT received $2 million in funding from the U.S. Department of Commerce to ensure competitiveness worldwide, protect national security, and develop engineering talent. Edwards noted that where funding used to come primarily from industry, there is now more investment coming from government.

A key government partner is the U.S. Department of Defense (DoD). ASU partnered with the DoD on a workshop to improve semiconductor research and development and training across Arizona. Then ASU and the Arizona Board of Regents were one of eight successful bidders to win funding through the DoD’s Microelectronics Commons regional hub scheme. The program aims to create lab-to-fab pathways to commercialization for U.S. microelectronics researchers and designers.

“The DoD thinks, for good and obvious reasons, from the perspective of national security,” said ASU’s Squires. “If they are vulnerable in some of their platforms or techniques, that exposes them. And we’re seeking an intersection between our expertise, the directions we want to go, and what some of those key platform technologies are. So it’s sharing ideas and saying, ‘You have this need, we have this expertise. Do we have complementary synergies?’ It turns out that we do, and this is a synergy that could provide funding to take it further. That will be important, because that will go all the way from discovery on the tabletop within the laboratory, to actual prototype in a DoD system. That’s the whole entire value chain.”

Academia/industry partnerships and funding from the CHIPS ACT can fuel U.S. semiconductor innovation in terms of facilities, equipment, and expertise, according to Squires. “While companies like TSMC and Intel have amazing scale, they need this neutral ground, where they could work with innovators, with small companies, with faculty, and begin to really understand if these innovations have a chance to achieve scale.”

One such partnership involves ASU, Applied Materials, and the Arizona Commerce Authority. The goal is to create a shared research, development, and prototyping facility to accelerate the transfer of innovations from ideation to fab prototype.

“The Applied Materials partnership began a couple of years ago where, ultimately, it’s a contact between some of their engineers, their leadership, and some of our faculty,” said Squires. “And it’s initially just mutual exploration. Phoenix is a focus for semiconductor manufacturing and innovation, so it makes sense that Applied Materials wants to be here. Their customers are here. Intel is here. TSMC is here. And it really does start that way. We have these problems, these needs, and you all have this expertise. And through those conversations, we begin to build small partnerships around projects, engaging students on things that the universities are good at.”

Meanwhile, universities must remain sensitive to geopolitical concerns. In some cases, participation may be limited to citizenship or green card holders. Security plays a key role in many of these relationships.

In addition, Squires pointed to issues around critical materials for semiconductor manufacturing. “Those are still coming from challenging parts of the world and that’s a whole other conversation that we’re beginning to build around,” he said. “So we ask, how do you think about reuse? How do you think about recycling? A lot of what we’re talking about is literally mining minerals from the ground. The way that has been practiced in the past is almost appalling. It doesn’t have to be that way. So there are a lot of things that will change, favorably, if we really get this right.”

Geopolitics adds another level of urgency to solving these issues. “I’m going to call it a crisis at this point,” said Cornell’s Xing. “First, we got impacted because of the geopolitical tension right now. We have a hard time getting the talent from certain countries right now. Some international students are facing more legal barriers to come to us. In that sense, in terms of the talent pool, we are definitely impacted.”

International collaboration
A number of partnerships formed between universities and organizations from different countries, with geopolitical issues only a background concern. For example, Micron and Tokyo Electron partnered with the U.S. and Japanese governments and 11 universities, including RIT, to cultivate a more robust and highly skilled semiconductor workforce for the two countries.

As for the Netherlands-Taiwan partnership, TU/e’s international relations director is from Taiwan and she pushed the partnership idea. “It was a natural collaboration, especially because we had a Taiwanese gateway,” said TU/e’s Todri-Sania. “We stayed away from politics, which is less of a challenge for academia than it is for industry. The partnership is about science. We as a university are very neutral though we do perceive a bigger push on the companies, and eventually that will affect universities.”

For the summer school, 15 students from TU/e and 15 students from Taiwan came together on the TU/e campus and there may be a return visit where TU/u students visit Taiwanese universities in the future.

TU/e’s Heck said he has mixed feelings about academic/industry partnerships because universities must retain “maximum freedom” to shape their curriculum and pursue research. However, provided that boundary is in place, Heck said partnerships can help in three key ways. First, they support activities and facilities, such as ASML’s recent funding of a state-of-the-art clean room at TU/e. Second, “We need to keep curriculum relevant. Physics and maths are basic courses, and we need people to be educated in those fields. But electrical engineering tech moves forward extremely fast. It’s the most advanced tech ever in humankind. If the tech is doubling two times every two years, the curriculum cannot stay the same. We need to stay on top of that and that’s why we need to link with industry.”

Finally, a great curriculum and partnerships with recognized organizations helps attract students and faculty.

Technology, security, and AI
To accelerate U.S. advances in information and communications technologies, SRC and the Defense Advanced Research Projects Agency (DARPA) launched the Joint University Microelectronics Program 2.0 (JUMP 2.0). Six U.S. universities were chosen to build and run a center dedicated to a particular high-risk, high-payoff area of research, ranging from next-generation AI to emerging memory devices and novel electric and photonic interconnect fabrics.

Cornell University won the bid to build the SUPREME Center for Superior Energy-Efficient Materials and Devices. Xing said a key goal is to develop different research methodologies and make them more efficient by leveraging artificial intelligence while not compromising on accuracy. Faculty worked together to decide which JUMP area Cornell should apply for

“Cornell has almost a decade-long history in advanced materials, so it is one of our most recognized strengths,” she said. “You want to build on your reputation, because reputation matters in terms of attracting your team, and establishing your workforce development, as well. So that’s a big angle. That’s one of the main reasons a lot of these partnerships have been formed. For the best talent, you always have to actively recruit. I believe the only area that does not need to actively recruit in the past 20 years is computer science.”

AI technology and cybersecurity are hot-button topics, leading to numerous university/industry collaborations, such as RIT’s partnership with Alstom to advance cybersecurity education, development, and research in the transportation industry, and another partnership between RIT and the Diplomatic Security Service to identify and address the latest techniques employed by malicious cyber actors.

“Cybersecurity is going to become more and more of an issue in everything related to manufacturing, the Internet of Things, or just how systems talk to each other,” said RIT’s Edwards. “It’s huge right now, and we are seeing that information being addressed by lots of different students and a lot of different majors.”

As for AI, Edwards said RIT has a training grant from NSF. “We’re covering everything from developing the technologies that will advance AI to using AI. And then we have a lot of researchers who use it for totally different things, whether it’s discovery of materials properties to health care.”

While both AI and cybersecurity grew out of computer science, they’re now crossing traditional boundaries between electrical engineering and computer science and computer engineering, both in terms of faculty research and curricula. “It’s important, but it also gives a student employable skills,” said ASU’s Squires. “The next wave will be much steeper in AI. And here again, the experts are already beginning to embed that into curriculum. But it will be a much broader initiative, which for ASU over the coming year will be really significant.”

Still, when it comes to hardware security, the challenge is getting students interested in the subject when there are few, if any, formal programs. So MITRE created an Embedded Capture the Flag (eCTF) competition among college students, and more recently high school students, to create a secure system for them to learn from their mistakes. The competition has since shown up other weaknesses in the education system, which is that semiconductor security also has a global supply chain element.

“In 2022, we had that as part of the competition, where the underlying hardware could be compromised,” said MITRE’s Walters. “We asked students to design their system to be resilient to a potentially malicious hardware component built right into their system. We got a really interesting response to that. A lot of students asked, ‘What are you talking about? How is this even possible?’ And the response was, ‘Yeah, it’s hard.’ But that’s what we’re doing in the real world. So you can either bury your head in the sand, or you can start working toward it and changing your mindset for how you design a system to be resilient. And then you’re going to be ready for it when you get out in the workforce, because that’s what your employer is going to ask you to think about.”

What’s interesting here is the competition is engaging students to think about embedded security even if it isn’t offered in the current curriculum. “They’re driving it, and they’re pulling in freshmen, teaching them what they need to learn, pulling in external speakers to talk at their weekly meetings,” he said. “They’re competitive, and the gamification of it makes them want to win. And they’re trying to teach the younger students and going out and finding resources, collectively making the whole group stronger so they can win the competition. But they’re also turning themselves into very valuable assets in the workforce for embedded security.”

Data
So who is partnering with whom? The following table lists prominent new partnerships between academia, the chip industry, and government announced in 2023, but there are many more beyond this list. The table is currently presented according to country, but it can also be sorted by university or industry/government partner.

Partnerships Table

Ed Sperling contributed to this article.

Related Reading
Rethinking Engineering Education In The U.S.
Academia, industry partnerships ramp to entice undergrads into hardware engineering.
Where All The Semiconductor Investments Are Going
Details on more than $500B in new investments by nearly 50 companies; what’s behind the expansion frenzy, why now, and challenges ahead.