Photon Bridge raises €10m to build laser sources for AI data centres at semiconductor scale
Photon Bridge has raised €10m co-led by Amadeus APEX Technology Fund and 360 Capital to scale laser sources for AI data centres. Samples due 2027.
Photon Bridge, a photonics company based in Eindhoven, announced a €10 million round on 21 October 2026. The Amadeus APEX Technology Fund, which APEX Ventures manages together with Amadeus Capital Partners, and 360 Capital co-led as new investors. Three existing backers increased their stakes: Innovation Industries, Faber and the Brabant Development Agency, known as BOM. None of the reporting gives the round a stage label, so it is best read as a €10 million raise and nothing more. Paul Marchal is chief executive.
The company builds laser light sources for the optical links that carry data between chips inside artificial intelligence clusters. Its approach is to combine and multiplex the light of many lower power indium phosphide semiconductor lasers rather than rely on fewer, more powerful ones. Lower power density per laser makes each one less sensitive to manufacturing defects, which the company says lifts yield. The platform targets up to 40 milliwatts of optical power per wavelength per fibre at high wavelength counts, and is designed to be produced across several indium phosphide foundries rather than at a single supplier.
The demand case rests on how quickly copper is running out of road inside large training and inference clusters. Omdia forecasts more than 20 million data centre artificial intelligence accelerators shipping each year by 2028, and the clusters themselves are moving from thousands of processors to hundreds of thousands. That pushes the optical link from the edge of the rack to a position beside the chip, in what the industry calls co-packaged optics, and every one of those links needs a laser. “The industry will soon need millions of these laser sources every year,” Marchal said. Ion Hauer, investment partner at APEX Ventures, framed the thesis more bluntly: “Every new wave of compute creates new bottlenecks further down the supply chain, and lasers are the next one.”
The money goes to product development, customer qualification and a larger team in Eindhoven. First samples are due to reach customers in 2027, with volume manufacturing from 2028, which puts roughly two years between this round and production revenue. IO+ reports that part of the manufacturing argument is mechanical rather than optical: a small flexible connection behaves like a tiny spring and lets the laser chip settle into position by itself, so standard assembly equipment can do work that would otherwise need nanometre precision robots. Cesare Maifredi, partner at 360 Capital, put the commercial shift this way: “AI is turning lasers from a telecom component into a semiconductor-scale product.”
Why it matters for AgTech founders: this is a component play inside somebody else's capital expenditure boom, and the structure of it is worth copying rather than admiring. Three things stand out. First, the moat on offer is manufacturing yield, not peak performance. Photon Bridge is not claiming the best laser, it is claiming the one that can be built a million times, and that is the argument an ag hardware founder should be making about a sensor or an actuator once the field trial is behind them. Second, designing for several foundries instead of one is a commercial decision dressed as an engineering one. Ag hardware teams routinely build a product around a single supplier's part and then discover, at the first volume order, that their gross margin and their delivery date belong to that supplier. Third, look at the capital shape: €10 million to cover samples in 2027 and volume in 2028 is a deliberately funded gap, not an accident. If your own machine roadmap has a gap of that length and no named investor who has agreed to sit through it, the gap is the risk, not the technology.
Photon Bridge raises €10m to build laser sources for AI data centres at semiconductor scale
The Eindhoven photonics company is backed by new investors Amadeus APEX Technology Fund and 360 Capital, with Innovation Industries, Faber and BOM increasing their stakes.
Photon Bridge, a photonics company based in Eindhoven, announced a €10 million round on 21 October 2026. The Amadeus APEX Technology Fund, which APEX Ventures manages together with Amadeus Capital Partners, and 360 Capital co-led as new investors. Three existing backers increased their stakes: Innovation Industries, Faber and the Brabant Development Agency, known as BOM. None of the reporting gives the round a stage label, so it is best read as a €10 million raise and nothing more. Paul Marchal is chief executive.
The company builds laser light sources for the optical links that carry data between chips inside artificial intelligence clusters. Its approach is to combine and multiplex the light of many lower power indium phosphide semiconductor lasers rather than rely on fewer, more powerful ones. Lower power density per laser makes each one less sensitive to manufacturing defects, which the company says lifts yield. The platform targets up to 40 milliwatts of optical power per wavelength per fibre at high wavelength counts, and is designed to be produced across several indium phosphide foundries rather than at a single supplier.
The demand case rests on how quickly copper is running out of road inside large training and inference clusters. Omdia forecasts more than 20 million data centre artificial intelligence accelerators shipping each year by 2028, and the clusters themselves are moving from thousands of processors to hundreds of thousands. That pushes the optical link from the edge of the rack to a position beside the chip, in what the industry calls co-packaged optics, and every one of those links needs a laser. “The industry will soon need millions of these laser sources every year,” Marchal said. Ion Hauer, investment partner at APEX Ventures, framed the thesis more bluntly: “Every new wave of compute creates new bottlenecks further down the supply chain, and lasers are the next one.”
The money goes to product development, customer qualification and a larger team in Eindhoven. First samples are due to reach customers in 2027, with volume manufacturing from 2028, which puts roughly two years between this round and production revenue. IO+ reports that part of the manufacturing argument is mechanical rather than optical: a small flexible connection behaves like a tiny spring and lets the laser chip settle into position by itself, so standard assembly equipment can do work that would otherwise need nanometre precision robots. Cesare Maifredi, partner at 360 Capital, put the commercial shift this way: “AI is turning lasers from a telecom component into a semiconductor-scale product.”
Why it matters for AgTech founders: this is a component play inside somebody else's capital expenditure boom, and the structure of it is worth copying rather than admiring. Three things stand out. First, the moat on offer is manufacturing yield, not peak performance. Photon Bridge is not claiming the best laser, it is claiming the one that can be built a million times, and that is the argument an ag hardware founder should be making about a sensor or an actuator once the field trial is behind them. Second, designing for several foundries instead of one is a commercial decision dressed as an engineering one. Ag hardware teams routinely build a product around a single supplier's part and then discover, at the first volume order, that their gross margin and their delivery date belong to that supplier. Third, look at the capital shape: €10 million to cover samples in 2027 and volume in 2028 is a deliberately funded gap, not an accident. If your own machine roadmap has a gap of that length and no named investor who has agreed to sit through it, the gap is the risk, not the technology.
