Photonic integrated circuits (PICs) technology is sparking transformations in applications as diverse as LiDAR systems, agri-food, and hyperscale data centers. But manufacturing PICs at scale demands a level of complexity beyond the scope of any single company. It will take an ecosystem. That is the principle behind PhotonDelta’s alliance of materials foundries, PIC designers, chip packaging firms, and other technology partners working together to fulfill the promise of PICs technology. PhotonDelta’s Eelko Brinkoff and Jorn Smeets join Dan McCarthy on this episode of Manufacturing Matters to explore the revolutionary potential and practical challenges of bringing PICs to market.

Dan McCarthy: [00:00:06] Hello everyone. Welcome back to Manufacturing Matters podcast, where we talk about the global trends and technology affecting global manufacturing. I’m Dan McCarthy of Tech B2B Marketing, recording here on-site at Photonics West 2025. Gratitude to SPIE for letting us participate here. And joining me today is Eelko Brinkhoff, CEO of PhotonDelta, and Jorn Smeets, the managing director of North America for PhotonDelta. We’re here to talk about PICs today, photonic integrated circuits. For those of you who have never heard of the technology, it’s like integrated circuits only with photons. Eelko, can you kind of kick off a little bit on what PhotonDelta’s mission is within that realm and maybe touch upon what PICs are, the benefits.

Eelko Brinkhoff: [00:00:47] Okay, first I’ll talk about PICs. So you said that we use photons instead of electrons. And the advantage is that you have faster speed and bandwidth. So a high capacity and lower energy consumption. You can use measurement and detection. So you see in the applications in datacom, high-performance computing, but also in the sensing part, that it offers a lot of new advantages and also new solutions that you can bring to the market. And you can also combine multiple optical systems on the chip and therefore make devices smaller.

Dan McCarthy: [00:01:23] There is a lot of potential there. And PhotonDelta’s got a direct role in contributing that technology forward.

Eelko Brinkhoff: [00:01:29] So PhotonDelta is a nonprofit organization. We’re a foundation, primarily publicly funded. And we try to be a growth accelerator for the photonic chip industry, specifically in the Netherlands but also more and more in the European level, where what we try to do is we try to fund startups, we have innovation programs, and we look at the challenges in the value chain as well as look to new applications. We do internationalization, so help our companies and our ecosystem to go abroad to find new business partners and technology partners. We do a talent attraction program. Because we’re going to grow this industry and we need the right talent.

Dan McCarthy: [00:02:06] On the workforce.

Eelko Brinkhoff: [00:02:07] Yeah, on the workforce. And we also do a technology road map. So what’s happening in technology, what’s happening in the market. And what impact does it have on which companies we need to fund, what innovation programs we need to run, which countries to focus on.

Dan McCarthy: [00:02:22] Right.

Eelko Brinkhoff: [00:02:22] And then, of course, we have somebody in the United States.

Dan McCarthy: [00:02:25] That’s right, Jorn Smeets. So really quickly, before we get into that, tell me a little bit more about the constituency of PhotonDelta. How many companies do you represent? Is it all startups? Is it also academic groups?

Jorn Smeets: [00:02:37] Yeah. In the community, it started from universities and knowledge institutes, and then we got startups. Some are now scale-ups. So if you look to what we talk about as our ecosystem, we have 70 companies involved in our ecosystem. So we come from four, and let’s say six or seven years ago, so now to 70. So we see about 10-12 new startups in the past year that we also funded. But we also see foreign companies that want to become part of the ecosystem or Dutch companies that want to diversify their business and also want to step into the tech industry.

Dan McCarthy: [00:03:13] You described it as an ecosystem just now, which is very specific. Tell me a little bit about what that means. It’s not exactly a value chain, but it is collaborative.

Jorn Smeets: [00:03:23] Yeah, it is. Of course, you can first look at the cluster. So you see the saturation. There’s a cluster of companies. So in the Netherlands you have two clusters that are quite dominant, one in the south and one in the east. One is working on indium phosphide. The other is silicon nitride. And you see there’s an interdependency in the value chain. So now this is quite unique because we have the complete value chain in both platforms, from design to foundry testing, packaging to application. And then you also have tech partners supplying equipment, machines. You have the universities and the knowledge institutes.

Dan McCarthy: [00:03:56] Okay. And you mentioned indium phosphide and silicon nitride. Is there also silicon photonics? Is that part of the mix as well?

Eelko Brinkhoff: [00:04:04] Yeah, the silicon photonics. We don’t have the foundry in the Netherlands, but we have a close cooperation with Imec in Belgium. And they have also silicon foundry capacity.

Dan McCarthy: [00:04:13] And so we’re talking about technology. Let’s talk about the applications where this is really going to drive some of the benefits. I know there’s three major groups. You have data centers, lidar, and biological sensors, medical sensors. I don’t know if any others have emerged since I’ve read up on the industry, but  where do you see most of the growth and most of the activity?

Eelko Brinkhoff: [00:04:34] Well, I think it’s happening. It’s happening now in datacom. You know, we’re reaching the limits. I think energy is limiting the growth of the semiconductor.

Dan McCarthy: [00:04:44] Conventional systems.

Eelko Brinkhoff: [00:04:45] Energy consumption. So you have to do something within the current footprint to bring the energy consumption low or keep it the same and then grow with the data capacity and the speed. Photonic chips could be a solution there. But in the sensor part, you mentioned lidar already. It still will take some more time, but also seeing agrifood. So you can, say, scan the land but also the freshness. And in health care, the sensing, you’re talking about bio-sensing. But also a photonic chip can do, 1,000th of a degree in temperature change, it can measure. So we have a startup that if you put liquid fluid in the blood vein, they can measure the conditions of the heart from the outside instead of you see it when it’s open.

Dan McCarthy: [00:05:42] Okay. And just to dive a little more deeply in the technology, I’ll ask you, Jorn, tell us a little bit more on the the data center technology. I know that it’s lower energy. It’s higher bandwidth. But what is it about photons that allows that that electron can’t do?

Jorn Smeets: [00:05:58] So basically if you look at the photonics industry, it’s not new to the datacom world. Optical fibers have been there for a long time. But the trend that we see is that the optics is migrating more towards the racks, into the racks, onto the chip. And we have near=packaged optics, co-packaged optics. So we see that it has benefits because photons are much faster. They have less resistance. And so especially when you look at the transceivers, the optical transceivers, which are the pluggables, those are migrating towards integrated photonics. Because when you integrate both the optical and the electronic components onto a chip, you have less loss. You can reach higher bandwidths, which is all crucial when it comes to the hardware requirements now for the AI era. But there are also other types of applications coming around the corner. Nowadays you see a lot of attention to optical circuit switching, OCS. A lot of companies doing that. You see some companies working on optical compute, like Q.ANT in Germany. And then there is of course the interconnects. That’s a very important one. Last Monday we had our event in Sunnyvale, the Photonics Innovation Day. And there again Nvidia was stressing the fact that the interconnect is the bottleneck right now for their GPUs to talk to the memory.

Dan McCarthy: [00:07:28] And that’s all on chip, correct?

Jorn Smeets: [00:07:30] Yeah exactly. Yeah. And so now that’s why you see companies like Lightmatter, like Ayar Labs raising massive amounts of money because there’s a lot of traction for these type of technologies. So that’s I think what we see now in datacom.

Dan McCarthy: [00:07:44] The way I’ve heard it described, Nvidia just wants to get the data off the chip faster. Right now it’s all done through connections, electro connections that have to be converted to optical a little further downstream. But they’re trying to do it on chip. And that’s what PICs can do.

Jorn Smeets: [00:07:58] Exactly. Straight into the optical domain. Because then you’re going to save a lot of energy. But also you can have those speeds that you need.

Dan McCarthy: [00:08:06] Right. And energy’s made the headlines, particularly AI. I know AI is a big driver for all of this as well because it’s very compute intensive. So now to steer a little bit, this is all wonderful, benefits. I know that there are some challenges to manufacturing PICs, particularly scaling up into volume, like it’s a semiconductor technology. So it’s all about yields. It’s all about volume. Let’s talk a little bit about that. What are some of the challenges that people are working on? I want to emphasize that there’s many, many solutions being explored out there. But what are some of the solutions that people are exploring to kind of scale up PICs?

Jorn Smeets: [00:08:40] Well, so I think when you look at the challenges in scaling up the production, one thing that comes to mind is the integration of the different platforms. I think Eelko described it nicely that there’s various platforms. We in the Netherlands are focused on the indium phosphide and silicon nitride. And then there is the silicon photonics that we work with Imec for. But in the end, there’s no single platform that is going to be able to address all of the requirements for all the various applications. So it’s going to be a mix and match depending on what you need as a customer, because in the end, the customer doesn’t care what type of material is in there.

Dan McCarthy: [00:09:15] We’re talking about wafer level here.

Jorn Smeets: [00:09:17] Yeah, exactly. On the wafer. Yeah.

Dan McCarthy: [00:09:18] So they don’t care what’s on the wafer. They care what’s on the chip.

Jorn Smeets: [00:09:19] Yeah, what does it do for me? And there we see some challenges with scaling up, because you need to integrate those materials. You know, they have different properties. How do you do it in an efficient way? And because also now standards is something that we’re lacking. So everyone is doing it in their own way. When you look at the packaging side, so I think around 70% of cost is going into the packaging. And if we can help to standardize some of those processes, then I think we’re going to make some big steps there. And one of the things that we do at PhotonDelta is we also in the past have funded some initiatives to develop specific equipment, for example to help streamline those processes. One of the fundings we did is also in MicroAlign, it’s fiber optic alignment, which of course as you know is an issue.

Dan McCarthy: [00:10:09] At the packaging level.

Jorn Smeets: [00:10:11] Yeah, exactly. And then you have the testing, wafer-level testing. It’s not really there yet. So PhotonDelta, we’re really stimulating projects and bringing together the organizations that have the potential to solve these type of issues.

Dan McCarthy: [00:10:27] Excellent.

Eelko Brinkhoff: [00:10:29] Maybe one thing to add also, we’re building up the capabilities and you’re building up the capacity. But in order to do so, sometimes you also need to create volume. And volume to understand the process, to improve the yield, to shorten the process, but it’s a little bit of a chicken and egg situation. Especially in datacom, it’s very simple. What you have? Is it reliable? Have you tested it? Did you already supply it to a customer? You know, we’re bringing new technology into the market. And we had the Photonics Innovation Day. If you listen to the story: Coherent, Nvidia, Lightmatter, they don’t necessarily agree on what direction it goes with what platforms.

Dan McCarthy: [00:11:11] Right. That’s right. It’s a little bit that you need the volume to drive the application, but you need the application to really drive the volume. It’s a little cart and horse problem. Let me back up for a second. This reminds me a little bit of the whole semiconductor industry back around the ’80s I would say. They had all these different players. Not a lot of standardization. Everybody was kind of working on their own thing. And it wasn’t until they really started collaborating on certain things that that industry took off. So we’re sort of back there again. But one of the distinguishing characteristics is that you’re using so many different materials. You’re just not going to escape from that because silicon doesn’t lase. So you have to use things like indium phosphate. You want to use silicon nitride for waveguides. The supply chain then becomes more complex. Because you’re not just dealing with one material; you’re dealing with any number of materials. How does PhotonDelta, do you have any contributing role in that realm?

Eelko Brinkhoff: [00:12:07] Well in the innovation program we run, we’re running a program of €1.1 billion that started in the beginning of 2023: €470 million by the public, by the Ministry of Economic Affairs, and €630 million by private companies. The consortium that puts in the innovation projects. We run 76 innovation projects. It’s also about heterogeneous integration and solving some of the challenges that we see in the value chain in maturing technology. And heterogeneous integration is definitely one. And you can look at the material but also look at the equipment that can make it happen. And also there you see different technologies, micro-transfer printing but also last Monday night I was at LIFT transfer?

Jorn Smeets: [00:13:00] Laser assisted.

Eelko Brinkhoff: [00:13:02] So also there we need to mature. A lot of things are happening. A lot of good things are happening. But I’m pretty sure in five years we’ll look back . . .

Dan McCarthy: [00:13:14] Some things have changed. Then we’ll talk again.

Jorn Smeets: [00:13:16] And I think maybe also to add to that is what you were saying, the comparison with the semiconductor industry back in maybe the ’80s, where we go from vertically integrated organizations more towards horizontally integrated organizations. So we have more specialized companies that have their role to play within the supply chain. And we see that same trend happening with integrated photonics now. And as PhotonDelta, we’re trying to bring in those companies that have their specific capabilities and have a role to play and that they know how to find each other and that maybe they work on common projects that PhotoDelta coordinates. Recently we added a Soitec, for example, to the mix. We’ve been adding some more design companies to the mix, like also GDSFactory, for example, Luceda. Really in order to make the supply chain as accessible as possible to the industry to start prototyping but then eventually also scaling up that production.

Dan McCarthy: [00:14:14] Okay. And that’s also part of your mission here in the U.S. is you’re trying to make those connections here because a lot of the design base is here. I don’t know how manufacturing is at this point, but there’s that and there’s foundries. GlobalFoundries, for example, has a photonics program. So we just talked about, there might be a very different conversation five years from now. I’ve been talking about this industry for three or four years, and it seems to be accelerating. Do you think that acceleration is going to accelerate further. Do you think we’re on a steady growth stage toward the heterogeneous integration and implementation on a broader scale?

Eelko Brinkhoff: [00:14:52] I believe it will, and it all started with technology coming from universities. So you’re pushing something into the market, but now you actually see the pull. So the pull will pull other companies in. The search for solutions is more fierce, especially in datacom. The energy problem in the U.S. is different from Europe. But in Europe, in the Netherlands, if you want to plug in a new company, it’s quite challenging to get the energy connection. So in Ireland, 70% of energy consumption nowadays is in data centers. If nothing happens in 10 years, there will still be 70%. So we cannot sustain that. So you really see a pull looking for the solutions. At the same time in maturing the technology, maturing the PIC itself, you see more and more startups in the application side, and if you look to the growth of our ecosystem in the past years, it’s especially startups in the sensing. So not necessarily in datacom. And really looking for applications in mobility, in the agrifood, in the health care, what sensing could add of value in it. And there you see really also new applications that nobody thought was possible five years ago. And that dynamic is increasing. And I think also recently we see that more foreign companies also want to be part of our ecosystem. So I think five years ago, three or four years ago, we were still preaching: “Look at this great technology.” We don’t need to preach anymore. And now it’s interaction. See what benefits the collaboration. What are the challenges, what are the needs, and how how can we help? And accelerating this industry to overcome the challenges or to support a collaboration?

Dan McCarthy: [00:16:37] Okay.

Jorn Smeets: [00:16:38] And I think you also see announcements, like this week TSMC announced their collaboration with Nvidia. There are some major companies stepping into silicon photonics to integrate the photonics, and that’s a testament of the momentum that’s there.

Dan McCarthy: [00:16:51] That’s a good point.

Jorn Smeets: [00:16:52] We’re convinced also now with GlobalFoundries, with Tower, all of these big guys stepping in the game, it means this industry is taking off. And we see the whole ecosystem of suppliers that are contributing to that. Also, riding on that wave of acceleration. So I definitely agree with what Eelko said. And you know, if you look at it from a system level, the complexity is increasing. Moore’s law is slowing. Integrated photonics is there to stay, and it’s only going to increase in volume as we’re addressing more on the sensing applications. There the volumes are far more important than in the datacom world. Looking at consumer applications. So I think we’re going to see some really cool stuff happening soon.

Dan McCarthy: [00:17:41] I agree. I’m looking forward to every bit of it. Is there anything else we missed about PhotonDelta that I didn’t share?

Eelko Brinkhoff: [00:17:47] We’re an organization of 30 people. We do the funding of startups. We did 10 startups in 2024. Two that we’re still talking to. The innovation program. So we run 76 innovation programs. The majority of the funding that we receive is in innovation projects, and we can subsidize up to 50%. And the consortium partners need to bring in the rest. So that’s also leveraging the public funding, really multiply by two and even more. And we hope to further increase that.

Dan McCarthy: [00:18:22] Excellent.

Jorn Smeets: [00:18:22] Maybe one more thing that I would like to add is on the applications. So we see an increase in application-oriented companies that are fabulous, that make use of that supply chain. And we at PhotonDelta that we try to stimulate the creation of new startups, new applications. So that’s why the end of last year we launched a engineering contest, a global engineering contest for people to submit their ideas on innovative applications that work with integrated photonics and that contribute to solving societal challenges. So far we’ve received a tremendous amount of applications. I think some of them are really cool, and we still have until the end of March to submit your ideas. So if you have any good ideas, please submit them. You can go to our website, PhotonDelta.com. In the search bar, type in “Global Engineering Contest.” Even if you Google it, you’ll find it. So yeah, we’re seeing a lot of cool stuff happening there.

Eelko Brinkhoff: [00:19:23] Maybe one thing to also add, of course, our funding is from the Netherlands and the Netherlands government, and more and more we’re also looking at the European level. And of course, you saw the EU Chips Act. In Pillar 1 they created these pilot projects where the first four are all semiconductors. But we lobbied together with a group of eight CEOs of integrated photonics in Europe to get also another pilot line on integrated photonics. And we managed to do that. So this summer, €380 million became available for the pilot line. And now we’ve got a third of that funding for building a new fab for indium phosphide from 4 inch to 6 inch.

Dan McCarthy: [00:20:13] When do you expect that to come online?

Jorn Smeets: [00:20:19] I think construction will start this year. Then hopefully it would take one, one and a half years to build it with clean rooms and then install the machines. So it’s a five-year project that you’re running. So I think in two or three years we have the operation running.

Dan McCarthy: [00:20:35] Wow, that’s a big deal. Indium phosphide. You can integrate almost any photonic function with indium phosphide, but the wafers are still small compared to silicon.

Eelko Brinkhoff: [00:20:45] So yeah. So it’s important for our ecosystem. It really also allows us to scale up I think also in the time that it’s needed, if I look at the current developments.

Dan McCarthy: [00:20:55] I look forward to hearing more about that. Maybe we can talk again next year or live or here back at Photonics West 2025. Thank you both for joining us. It’s been a really enlightening conversation. It’s always great talking to you guys. This is Dan McCarthy, again, recording on site at Photonics West 2025. This is the Manufacturing Matters podcast. If you have any questions for Eelko or Jorn, please send them to us. We will forward them on, and you can see this episode at Manufacturing-Matters.com anytime you’d like to see it. Thanks very much.