The world is currently playing chess with dependencies and monopolies. On that board, photonics is rapidly gaining value. Bringing the chain completely to Europe is out of reach, but influence is created by occupying strategic fields. Now is the time to invest in our strengths, so we won’t be outplayed by Chinese and American grandmasters.
– Data centres and AI are driving supply chain maturity.
– ‘You can talk endlessly about applications but in the end, it comes down to manufacturability.’
– ‘Investments should go directly to companies.’
– Specialisation in critical links yields more influence than complete chain control.
‘The industry is shifting, which offers opportunities for Europe’
Not everyone may have noticed, but photonics has been part of the digital infrastructure for years now. Light plays an increasingly important role in data centres, fibre optic networks and communication systems. ‘Without integrated photonics, we wouldn’t even be able to have this conversation’, says Ewit Roos, general partner at Photon Ventures, in the Teams call. ‘Integrated photonics has been used in transceivers for a long time; not using them is no longer an option. Photonics is not a pipe dream, developments are currently moving very fast. In the next ten years, photonics, alongside electronics, will become a central link between digital intelligence and the physical world.’
So the technology is there, operational and scalable. ‘Energy consumption is now under enormous pressure due to the influence of digitalisation and AI’, says Roos. ‘The way we traditionally transport data from A to B is faced with physical and energetic limits. This applies within a rack, between racks and between data centres.’ In this environment, photonics shifts from a supporting technology to a structural component of systems. ‘Integrated photonics offers a lot of bandwidth in data, but also in wavelengths. This means, for instance, that in photonics sensing, you can cover a broad spectrum, from blood saturation to blood sugar with the same device. Also, photonics is more energy efficient. Transporting a bit with photons costs less energy than via copper and it can also be done over longer distances.’
These advantages are nothing knew, but only become decisive when they can be deployed economically. ‘You can talk endlessly about applications but in the end, it comes down to manufacturability’, Roos explains. ‘A lot is possible on a laboratory level. We see impressive examples of that in medical diagnostics. But as soon as things move to production, costs, volumes and chains become decisive. AI and digitalisation act as an industrial driver. Like a pipe cleaner that forces the entire chain, from design to supply and production, to mature. This will allow applications in sensing, diagnostics and spectroscopy to be developed much faster.’
Shifting market
Johan Feenstra looks at the same development from the production side. The CEO of the Eindhoven photonics foundry SMART Photonics sees a sector that is in full swing. ‘The photonics industry is still very vertically organised at the moment’, he says. ‘Companies develop their own chips, build their own modules and sell them as end products themselves. This is similar to the electronics industry in the 1970s and 80s. But it’s slowly shifting. Parties are emerging that focus exclusively on design, packaging or chip production.’
SMART’s foundry model also fits into that context. ‘By building a strong, independent photonics chip factory, you offer many more companies access to the technology’, Feenstra explains. ‘Competitors don’t want to go to the fabs of vertically integrated players to have their chips made, because that means they’d have to share sensitive information. They don’t have that problem with an independent foundry like SMART.’
According to Feenstra, the shift in the industry offers opportunities for Europe to acquire a unique position. He believes this is the ideal time to invest. ‘Because the chain hasn’t crystallised yet, you’re at the beginning of a new development. You’re helping to build something that’s still taking shape. This is fundamentally different from playing catch-up in an existing sector, where Europe has been lagging behind the rest of the world for many years.’
Indispensable links
Roos believes scarcity is an important signal in that process. ‘Right now, there’s a shortage of lasers, modulators and amplifiers, all crucial photonics components. That’s because capacity is lagging behind and there’s enormous demand for energy-efficient, high-bandwidth technology. This scarcity shows that photonics is moving to the heart of systems’, he explains. Feenstra adds: ‘When something becomes scarce, that’s when dependencies become visible. This gives technology such as photonics geopolitical significance.’ He deliberately avoids the word autonomy because he believes Europe having to be able to organise everything itself is unrealistic. He prefers the term indispensability. ‘It’s about links that others can’t ignore. ASML is a good example of this. A position like that provides leverage, without having to control the entire chain yourself.’
When asked, Roos outlines the global photonics landscape. ‘There’s more diversity than in the production of electronic chips. Photonics fabs can be found in the United States, Europe and Taiwan. The Netherlands is a leader in advanced photonic integrated circuits (PICs) based on indium phosphide and silicon nitride. There are a couple of players in Taiwan and they’re growing very fast, but that’s about it.’ Roos welcomes this specialisation: ‘It’s not about doing everything yourself, but about strategically valuable positions at specific points in the global supply chain. Trying to do everything in Europe would be extremely foolish.’
Chips Act 2
This consideration affects the way in which Europe has organised its industrial policy. Feenstra points out that photonics has long been an implicit participant in broader semiconductor programmes. ‘When Chips Act 1 started, there was a strong emphasis on electronics’, he says. ‘It took a lot of lobbying in The Hague and Brussels to gradually get photonics included. Fortunately, photonics will be included in Chips Act 2 from the start, which is scheduled to be come into force on 1 January 2028.’ At the same time, Feenstra puts the importance of policy documents into perspective. ‘Providing direction is important, but it’s the execution that determines whether companies can actually invest and scale up.’
Nevertheless, Feenstra is positive about the direction Europe is taking with Chips Act 2. ‘It signifies a clear choice and long-term vision.’ At the same time, he is concerned about the shorter term. ‘The next two years will be decisive for companies that have to invest now. Specific relief in terms of regulations, energy costs, and financing is still lacking.’
More direct investments
Policy plays a decisive role in this. Feenstra is outspoken about that. ‘Investments should go directly to companies’, he believes, and he immediately links it to cutthroat regulations. ‘The OIM Strategic Plan, for instance, is a nightmare. It punishes scale-ups that simply have their financing in order. Because if you need to scale and require larger amounts while turnover is still lagging, you switch to non-equity to prevent dilution. But that means that suddenly, you no longer meet the requirements and the money goes to companies that have long been profitable; that’s just ridiculous. There’s some movement, but the strategic plan is still there. I hope the new minister will sort this out quickly.’
‘It’s about links that others can’t ignore’
The tax on energy also has to be lowered, Feenstra argues: ‘SMART is getting a larger cleanroom soon, one that requires a lot of energy. It’s going to be expensive, and because of the high energy prices here, we’re losing competitiveness.’ Energy prices and grid connections determine whether production in the Netherlands remains feasible. He also believes that Europe needs to handle state aid differently. ‘We’re competing with China, and I don’t think they know the meaning of that word. They call it “economic development”. And in the US, too, government money is invested directly in companies. That’s something we need to do as well.’
Courage
Feenstra realises that this means clear choices must be made. ‘The National Technology Strategy shows courage because it includes industries that we don’t actually have yet or that we’re building, such as photonics. Only, the problem with the NTS is that no money has been made available for it in the past two years.’ But Feenstra thinks creatively: ‘It costs absolutely nothing to give priority to companies in the NTS sectors if they need a new or larger electricity connection. I understand that you’re not happy if you’re not on that list, but making choices is the only way we can be competitive.’
He draws a parallel with South Korea in the 1990s: ‘The country was bankrupt, and the Korean government determined where the country wanted to excel and supported that for years. Samsung and LG in displays, Hyundai and Kia in cars. And look where they are now. They did something similar in Taiwan with electronics and chip production.’ For Feenstra, such choices are essential. ‘Because if you don’t make a choice, someone will eventually make that choice for you. Or as Peter Wennink recently put it: “If you don’t have a seat at the table, you’re probably on the menu.”’
Choose, don’t share
Roos agrees with Feenstra: ‘If you want to stand against the violence from the United States and China, you shouldn’t do the same; you must excel in something that is critical in their value chains. So Europe must focus on the areas where we can truly excel, where we can leave the competition far behind through continuous innovation. And truly choose. Because if you distribute the resources across many initiatives, everything gets diluted and you end up being good at nothing.’
He does have an idea about those links. ‘There’s no doubt that we’ve established a control point for the production of indium phosphide photonic chips with SMART. Thanks to Phix from Enschede, we have an iron in the fire when it comes to packaging. And Microalign from Eindhoven also plays an essential role in the chain to make quantum computing a reality with its ultra-efficient light coupling on a photonic chip. But as we see at companies like Google and Apple, in the end, most of the money is earned through system integration, and properly embedding the technology. So we will have to build a number of substantial European integrators, because they don’t exist yet.’
Feenstra adds: ‘Seeking consensus isn’t always the best way out of a crisis. So within Europe, we shouldn’t be chauvinistic and really look at our strengths. We shouldn’t be building indium phosphide foundries all over the place, but support other strengths in France, Spain, and Italy. Because if there was ever an opportunity, it’s now. And not choosing is losing. The world doesn’t wait.’
