The Next-Generation Talent Question

Today, an electronics engineering student in Europe will mostly learn on proprietary tools. That is what their professor has set up. That is what the university’s licensing arrangement covers, usually under an academic discount of 99 percent or more from one of the three big EDA vendors. And that is what students want to learn, of course, because that is what companies hire for.

Jason Cong, a computer scientist at UCLA, has argued for opening chip design to engineers trained in software. In a 2024 conference keynote on “democratizing” chip design, he described teaching undergraduates with no hardware background to design a working chip accelerator in about a week and a half. European academia is not yet structured around that insight; chip design programs tend to treat software as a tributary rather than as a peer discipline.

Noam reframes the choice from the student’s perspective: “If you want a job, learn proprietary tools.” Industrial recruitment is downstream of the tools on a graduate’s resume. “If you want to start a startup, learn open.” New companies cannot launch without paying a million-euro license bill before they have even built a prototype. These are two paths into the same industry, and the second path only exists thanks to the open ecosystem.

One more piece of the puzzle, from Luca: Many of the same professors who could add open tools to the curriculum have an interest in keeping their relationships with proprietary vendors. Discounts of 99 percent or more are contingent on vendor goodwill, so a professor who agitates publicly for open alternatives may find their next discount renegotiation harder. That dynamic is hard to see from outside academia, but it almost certainly shapes what gets taught.

This is the same pattern that locks European public administrations into Microsoft and Amazon Web Services at a different layer of the technology stack. The argument is not that the proprietary tools are bad: Having no real alternative is. Digital autonomy means having options available.

What Public Money Buys

The first EU Chips Act, passed in 2023, put roughly 43 billion euros behind one goal: doubling Europe’s share of global chip manufacturing by 2030. The proposal for the Chips Act 2.0, adopted this June, aims to position the EU within the broader ecosystem by complementing newer initiatives such as the Cloud and AI Development Act and the Open Source Strategy.

The goal of reducing dependency has sometimes seemed out of reach, and the European Court of Auditors has pointed to the gap between European research and European fabrication as the core weakness — the second version of the Act explicitly focuses on increasing demand and innovation. The open EDA community has a sharper critique. A meaningful share of that public money has gone straight back out the door as license fees, paid to the same three proprietary vendors the investment was meant to make Europe less dependent upon. European startups working inside EU-funded design centers build their chips on cloud-hosted Cadence and Synopsys: software the public paid for and the vendors still own.

EuroCDP, the EU Chips Design Platform, is one potential answer. Coordinated by the Belgian research institute imec with 12 European research partners, it was set to onboard its first startups and SMEs in early 2026. EuroCDP gives European chip designers access to commercial EDA at negotiated discounts and to open source EDA tools, plus fabrication, training, and startup support. Within the platform, open source is treated as a pillar in its own right, with the same standing as the commercial offering. Whether the open pillar grows in proportion to the commercial one is the big “open” question.

What makes open silicon possible today exists in part because the European Commission and member-state funders have backed it for years, through programs few Europeans know exist. Increasing the technological sovereignty Europe says it wants comes down to sustained investment of public money in building tools, rather than renting them.

The Chips Act 2.0 broadens the agenda to include packaging, materials, and cross-border investment in fabrication, and it runs directly into this choice. Money spent on licenses is money not spent on engineers, designs, or training, so the same public grant goes further when the toolchain underneath it is free and open — the same is true when it comes to industrial design and fabrication. The second wave of European semiconductor legislation introduces “Grand Challenges” and “Demand Accelerators” to offer tangible resource support at the policy level, bringing public procurement money into the open silicon space.

This article was made possible by support from NLnet Foundation through Linux New Media’s Topic Subsidy Program. NLnet Foundation is the coordinator of the NGI Zero Programs which are funded by the European Commission.

Links for further reading:

• EU Open Source Strategy Plays Key Role in Tech Sovereignty Package
• Google’s Open MPW shuttle program announcement
• OpenROAD Initiative News and Updates
• EuroCDP
• European Commission Policy Page: Chips Act 2.0
• European Parliament Legislative Train Schedule for tracking progress
• Open EDA ecosystem supported by NGI Zero
• NGI Zero Programs

Links for your toolkit:

• F-Si EDA toolchain directory (GitHub)
• F-Si EDA and chip-design thesis database
• SkyWater SKY130 open PDK (GitHub) 
• IHP Microelectronics open PDK (SG13G2) 
• Naja (GitHub)
• Kepler Formal (GitHub)