How Companies Onboard Graduates And Shape University Curricula

A shortage of senior engineers with the necessary skills and experience is forcing companies to hire and train fresh graduates, a more time-consuming process but one that allows them to rise through the ranks using the companies’ preferred technology and systems.

Universities and companies share the goal of helping a graduate become productive in the workplace as quickly as possible, and there are ways they can shape curricula to reach that goal more quickly. Among them:

• Provide resources and tools for universities to embed in their courses.
• Give guest lectures or supply real-world problems for projects.
• Map job descriptions into modules.
• Attend job fairs and community engagement events.
• Offer formalized input by sitting on advisory boards.
• Join a national or global education network to pool resources and ideas.

Companies source talent from multiple university degree programs and across all levels. For EDA and AI companies, electrical engineering (EE), electrical and computer engineering (ECE), computer science (CS), physics, and math are all relevant. A master’s in business administration (MBA) is also valuable, particularly for sales or operations. A Bachelor of Science (BS) is often sufficient, but a Master of Science (MS) is common. A doctorate (PhD) tends to be needed for more advanced research positions only.

Building junior talent vs. hiring experienced engineers
The chip industry’s talent shortage dates back several decades. For example, Siemens created its 12-month Associate Applications Engineer Program during the dot-com bubble in the 1990s, when companies faced hiring and retention challenges due to the proliferation of startups. Today’s engineering job market faces a range challenges, including an aging workforce, competition from popular AI and hyperscaler jobs, the cyclical nature of demand, and even pressures from government programs.

“Looking at the retirement cliff in our experienced workforce, we absolutely have to have a pipeline of new talent coming into the organization,” said Michael Philippi, technical program manager at Siemens EDA. “The best place to do that, from a pragmatic standpoint, is to make the investment and hire people out of university, train them, give all that institutional knowledge and culture, and what the level of sophistication our customers need. Then, when we promote them and put them in the field, they at a level where they can be effective. And if we lose the best of our cohort, we’ve got people coming behind them who are quickly learning.”

Others agree that training graduates offers the simplest way to fill the talent pipeline. Arm mostly builds its own talent, but it also will pay for a senior expert in certain fields.

“A really high percentage at Arm are graduates,” said Khaled Benkrid, director of education and research at Arm. “Many of our executives started at Arm as a graduate or intern even, and it’s because of our timelines. We are the R&D lab of the semiconductor business. Things that we’re working on today, some of it will not hit the markets until maybe 5, 10, years henceforth. For that we need a really sustainable flow of talent, and we need people to grow with us and learn through osmosis. We put more of an emphasis on developing more junior people. However, on the more advanced, bleeding-edge stuff, you have no choice but to recruit from the market. It is challenging, but Arm has its appeal, as well. So far, we haven’t had big issues.”

Smaller companies and startups may rely on hiring graduates because of the tight competition for senior talent. “We struggled to find more experienced hires in the very beginning, but now we have got a real mix of experienced hires and graduates,” said Monique Williams-Lesser, director of human resources and operations at Axiomise. “It is quite difficult to find new employees with 10 years of experience, and a lot of the time they don’t want to leave where they are. With graduates, you’re dealing with fresh eyes and new ideas that really help to bring about creativity.”

Axiomise also maintains a broad hiring base to avoid missing out on talent. “It really does benefit an organization to have a mix of cultures, gender, and experience,” said Williams-Lesser. “That’s been the winning formula for us.”

Meanwhile, AI companies are competing in the Wild West of a frontier technology. “In general, Expedera has the benefit of being in a very fast growing industry that a lot of people want to work in, but we have to balance it against the fact that we’re competing against much larger companies with much bigger and deeper pockets, not only for salaries, but also for what resources you have to join the company,” said Paul Karazuba, Expedera’s vice president of marketing.

Expedera relies on a scaling model when hiring talent. “Let’s say we’ve got two really good senior engineers in one office,” said Karazuba. “Then let’s bring on the juniors and train them. Whereas in another office, maybe we need to concentrate on hiring a couple of senior engineers before we move into new graduates. A junior could include a fresh graduate or a person who has been doing something else in semiconductors for a few years and is looking for a market change.”

In-house training and internships
Shortening the path to productivity is a key goal, so it’s essential a company finds tasks that a junior engineer can do right away. Online customer support tickets are useful in this respect because the trainee can work on a problem under guidance from their mentor without needing to navigate real-time, face-to-face interaction.

“We used to hire experienced talent and then they had to sink or swim,” said Siemens’ Philippi. “What we’re doing now is hiring people, training them, and investing in their success, so that when they graduate 12 months later, they’re at almost the same level of proficiency in terms of doing the customer-facing technical sales job as someone with three years’ of experience. The focus of the program is on day-to-day training on our solutions. We’re not relying on the universities to teach them really advanced topics. What we need the universities to do is teach the fundamentals exceptionally well. Then, the most important tool we use is our trainers. When they graduate from the program, they are an expert in a fairly narrow space of our solution portfolio in one specific application space.”

Other companies have similar programs to train employees. “The Ansys Learning Hub is an interactive learning platform with resources that span Ansys products and industry applications,” said Kelly Sullivan, senior manager organizational development at Ansys. “This is offered to employees who work closely with products or interact with customers.”

Hands-on training with mentors takes a larger role when technology is very new. “We formalize our internal training processes, but a lot of it is not necessarily material,” said Expedera’s Karazuba. “It’s literally someone sitting here looking over an engineering shoulder and questioning, ‘What is that in the code? Why are you doing this? Or why have you taken this approach?’ Any materials we use are ever-evolving. Interns will look at the company webinars and speeches at events. Then, it’s getting down to work alongside a mentor, whether it’s coding, programming, or design. And some of these interns make a difference from day one. One of our interns from last summer is named on one of our patents.”

Similarly, Axiomise relies on hands-on training. Graduates receive access to its formal verification modules, interactive learning platforms, technical documentation, and workshops. Each new hire gets a mentor for support on real world projects.

For many companies, training begins during its internship program and continues once a graduate is hired full time.

“We’re a little less than 70 people worldwide, and more than 10% of those are ex-interns who’ve turned full-time,” said Expedera’s Karazuba. “We had 12 this summer. We find the universities are teaching the basics really well. For the more focused aspect of what the jobs entail, it’s incumbent upon the companies to train these employees, and we found that internships are the absolute best way to hire.”

Internships also give companies the chance to try before they buy, said Arm’s Benkrid. “And usually people are transformed after that experience, especially on the technical aspect, but also the non-technical — the attitudes, the behavior, the maturity — after a short period of time.” The conversion rate to full-time employee is “quite high,” and then new recruits enter Arm’s two-year Graduate Development Program where they rotate around different groups and get exposure into various parts of engineering with support and mentorship, said Benkrid. “In that time, they work on real-world projects and are productive right away. At the end of the two years, they are ready to tackle any problem. They choose a particular area, and they specialize in it.”

Engineering basics vs. the bleeding edge
It is easier for a company to adjust its onboarding process than for a university to overhaul its curriculum, but neither can remain static in such a fast-changing industry.

“Most university curricula change very slowly and rarely match the pace of industrial changes,” said Franchon Warmack, talent acquisition programs lead at Cadence. “Professors rarely change curricula because of the ABET requirements, or because they do not have the time or funding to update and refresh their materials to match the changes in the industry.”

Arm’s Benkrid agrees the underlying problem is economic. “When you get into the bleeding edge of things, the cost of creating educational content or training content is prohibitive. It’s too big for universities. The other thing is that bleeding edge changes quickly. In the old model, which has worked very well for centuries, technology wasn’t moving at a very fast pace. As an ex-academic, the way we used to work is to do a major revamp every four or five years, and in between that time we freeze the requirement of the curriculum. We invest heavily in the first year to create the content, train people, etc., and then we rely on a constant and predictable supply of students over the 4 or 5 years, so we amortized the cost –– the initial cost of development –– over 4, 5, or 6, years. But if you are having to revise the content on a more regular basis, and you don’t have a predictable supply of learners, it becomes a very difficult proposition.”

Despite these challenges, whenever the definition of foundational knowledge shifts, the curriculum also must shift.

“How much of the load of equipping future engineers for the right jobs should be taken by the education system, and how much of it should be taken by industry? The answer is that you need a collaboration between the two,” said Arm’s Benkrid, a former professor of computer science at Queen’s University, and of electrical engineering at Edinburgh University. “And what is foundational knowledge and skills is changing, as well. You can’t just freeze those requirements and say, ‘Well, the education sector will look after that.’ For example, even [grade] schools have to teach AI during pre-university education, whereas that wasn’t a foundational thing in the past. Now it is foundational, because the way we program computers is changing fundamentally. It’s a constant negotiation and a constant collaboration between universities and industry, and the line will shift over time.”

Others agree that universities should continue teaching the basics very well, but AI is no longer optional. “It’s difficult for anyone to keep up in the AI industry,” said Expedera’s Karazuba. “But if AI is really to be this earth-shaking evolution of what engineering and tech is going to be, it becomes incumbent on any engineering program –– regardless of the discipline –– to teach the fundamentals of AI and machine learning, how it works. I would back it up to say that an introduction to AI for anything in the corporate world is going to be important in the future.”

Overall, Karazuba believes the quality of new graduates is excellent. “Certain people want to demonize young people,” he said. “I have the exact opposite opinion. The people that we’re getting out of university are insanely talented, good people who work hard, and we’re seeing a lot more of them over the last year as university programs have grown and adapted.”

Programs and tools to bridge the knowledge gap
Compared to overhauling a degree program, short courses, certificate programs, or massive open online courses (MOOCs) can be deployed quickly, with the cost of developing two or three iterations recouped over a year, according to Arm’s Benkrid.

“Many universities are unbundling content into micro-credentials, where they offer shorter courses that are more amenable to leading edge technologies or things that are not as foundational,” he said. “These provide flexible, agile ways for more industrial input into the education curriculum. MOOCs platforms are also really helpful because they expand the user base. With our online courses, we have 80,000 plus people registered to our content –– we would have never reached that scale otherwise –– and then you can revise them on a regular basis.”

Industry-academia curriculum partnerships offer advantages both ways. “By embedding curricula into universities, we’re getting a lot of recruits from the universities where they are teaching with our content, and the feedback has been great,” said Benkrid.

For example, Arm’s Education Kits are a curriculum-aligned product with lecture notes, lecture materials, lab materials, and solutions all available on GitHub for people to adapt to purpose. “Professors can use them ‘as is,’ and they typically fit into a 10- to 14-week term course,” said Benkrid. “Anglia Ruskin University is starting a master’s course on embedded systems and the core of that content is based on four courses that we created, and then they’re adding to it.”

Similarly, a university in Spain took one of Arm’s online courses and translated it into Spanish for the Latin American market and had the product unlocked within a few months. And King’s College in the UK is building a master’s course for VLSI SoC using Arm’s content as foundation.

Another offering is the Arm KSA (knowledge, skills, and abilities) Framework, which identifies what is needed for a particular role in the semiconductor industry. “We just published our first KSA competency framework for early career software engineer and hardware engineer,” said Benkrid. “It has huge applications, and we are working with partners to augment it and develop other KSAs for different job roles. We then feed those requirements into curricular development and share this with the whole Arm’s Semiconductor Education Alliance and beyond. It’s all on GitHub now.”

Cadence, Siemens, Synopsys and Ansys are all members of the alliance in addition to running their own programs. For example, “Ansys Education Resources are modular curriculum units that are designed to enhance existing courses and are often developed in partnership with educators,” said Susan Coleman, senior director of academic and startup programs at Ansys. “Innovation Courses are bite-sized courses for self-learning that are often assigned as homework to supplement in-classroom learning.” And the Ansys Academic program helps educators incorporate Ansys tools to prepare engineers for industry needs and provide a hands-on learning experience.

Meanwhile, more than 350 academic institutions are enrolled in the Cadence Academic Network, and the company has a one-year structured Technical Pipeline Program consisting of online and on-the-job training focused on chip design for interns and fresh college graduates. “Many companies are working closely with schools to augment their curricula so that the graduates are readily employable and shorten their time to contribute when hired,” said Warmack.

Others agree that targeted credentials offer a bridge between university curricula and industry needs. “But some people in academia may find that to be a distraction,” said Russ Klein, program director at Siemens EDA and adjunct professor at Portland State University. “It’s just the pendulum between how practical and how theoretical does the education process need to be?”

A simpler way to support a university is to provide tools. For example, Siemens donated two emulators to Portland State and then hired Klein when a position became vacant. “It’s about helping out the school and also getting people who are going to be moving into the workforce familiar with our products,” he said. “It often happens that the first product in any particular category that you use, you think of this as the right way to do things. And if we can get students using the very first logic simulator, or the very first synthesis tool, or the very first whatever from Siemens EDA, they think of that as the right way to do it. When they get out into the industry, they’ll be looking for opportunities to be able to use that.”

Guest lectures also can keep students up to date on technology and function as an introduction in the existing curriculum.

“We usually ask universities to add more flexible ways for interventions as new technologies and new skills come on board,” said Arm’s Benkrid. “Angela Ruskin University has Live Briefs in its normal curriculum courses, inviting industry partners to come in and provide a real problem they want to solve to a group of students to tackle over two to three months. That gives students a great insight into real-world problems and it provides industry access to talent and problem-solving capacity. For projects, challenge-based or problem-based learning packages that allow industry to intervene and provide input and guidance are also helpful.”

Benkrid also sits on an industry advisory board with Anglia Ruskin that meets once or twice a year to look at the curriculum holistically and identify ways to improve it. The link between industry employability is now a big key performance indicator (KPI) for universities, he said.

Conclusion
Each company is coming up with solutions to meet its hiring needs in a shifting landscape. Yet networking and sharing resources offers an easier way to fill the talent gap in a global supply chain.

Arm takes a holistic approach to the education system, because it’s all connected, right from schools all the way to research, said Benkrid. “The key to the Semiconductor Education Alliance is that we built it as a federated effort rather than a hierarchical organization. Working together is the way to go to address what is otherwise a very, very difficult problem — the talent shortage. It’s about how do we plug the education and skills gap. We’re doing it for the whole ecosystem, because we succeed when our partners succeed, and our technology is so foundational that the branding aspect of it matters less for us because a lot of it is hidden from the end customer. We want to train people for our silicon partners, or our OEM partners. It’s important for us to solve the problem for the whole ecosystem, rather than specifically for our market segment.”