As 2025 draws to a close, I wanted to take a moment to share some of my reflections and insights on this past year. It’s been an incredible ride for Oriole with enormous progress on both the engineering and commercial fronts. I’m immensely proud of how we executed as a team and how we’ve built out the company. Beyond that, we’ve had countless opportunities to engage with customers, understand their needs, and validate our vision. What is clear is that we are firmly on track to realizing that vision.
Rising Awareness of Photonics in AI and the Need for Radical Solutions
We are also seeing an incredible dynamic in the industry, with the opportunity and the scope of the opportunity continuing to grow. The market is coming towards us with the increasing recognition for the need for photonics in AI and a growing realization that radical solutions are needed to tackle a problem that’s only getting bigger.
If you look at energy projections for AI, it’s clear: we are going to need way more power. But doing things the way we do now won’t solve that challenge. It reminds me of that moment in Jaws when someone says, “We’re going to need a bigger boat.” Maybe we don’t need a bigger boat – we just need a better one. And we believe Oriole is that better boat. The idea of photons whizzing around the data center once seemed plausible but futuristic. We’re saying it’s here and coming much sooner than anyone expected. Things need to be done radically differently moving forward, and we are that radical solution.
Especially in 2025, many traditionally silicon-based players woke up to the need for photonics on their roadmap, and fast. Their sense of urgency is high, and they are searching for the solution. And we believe we have that solution. Another point of observation is that Oriole seems to be the only company developing a complete solution for a photonic network in AI. Some people will say that they have optical networking – optical pipes between electrical switches, but that’s not photonic networking, and that’s a distinction that matters.
Coming back to the increasing recognition for the need for photonics in AI, NVIDIA made a very smart move a few years ago when it bought Mellanox. They wanted to go beyond being a traditional GPU vendor, and they realized the importance of AI in their future. Mellanox was the fastest networking solution out there, and by acquiring them, NVIDIA was able to bring together compute and networking to build a complete co-developed, co-optimized solution.
What we are seeing now is that if you want to compete against NVIDIA as an xPU vendor, you need to figure out how to get a network that is generationally ahead. That is a key strategic piece of the puzzle that companies that traditionally made silicon compute chips need to get their hands on. But how do you do that? We see companies popping up all the time that are making optical widgets. Those widgets are important pieces, but what matters most is how you build the solution. All these widgets might or might not be what you need when you actually build the solution.
Building the Next-Generation AI Network
The right way to come at it for the next generation network is to figure out what the solution is and then figure out what pieces you need. That is what Oriole is doing. Our approach is rather unique: we didn’t set out to build a photonic widget or even solve how to build a photonic network. We set out to build the highest-performing AI network imaginable, pushing toward theoretical limits. We are building a solution from the ground up with photonics as a key enabler. PRISM is more than an OCS; it is more than a co-packaged optics optical pipe. It is a massively performant AI network based on incredibly efficient photonics.
That’s a big paradigm shift, and it takes people, even the experts, a while to wrap their heads around what we are doing. We are not following a known path, with a series of known steps, while driving incremental improvements in a familiar technology like electronic switching. Moore’s Law isn’t delivering as a solution to the problem. Therefore, we are saying: get off the beaten path, tear it up, and find a different, fundamentally better way of doing things. And we have. Oriole has a solution that achieves the fundamental performance limits AND dramatically cuts the need for power.
PRISM and Beyond
For us as an organization, 2025 has been all about executing against the key decisions we made early on in 2024. Seeing our vision for PRISM come together and working has been a highlight of this year and a proud moment. We nailed our milestones, and the market is moving toward us, validating our original thinking. I’m also proud of our team. What we are doing isn’t easy, but we are going as fast as humanly possible, supported by incredible investors and a truly world-class technical leadership team. These people have the capacity to execute and deliver solutions to market, and they’ve attracted exceptional talent to join us. I think we have amazing people, and I have not seen a better company out there.
Another highlight for me is that governments are waking up to the fact that this is a strategic industry for sovereignty. Governments need to take a bigger and more active role in supporting and funding this type of innovation. AI will impact everywhere and everything we do. Nations need to take control over their AI destiny, which is directly tied to the underlying network infrastructure.
Reflecting on 2025 also offers the opportunity to look forward to 2026. Our biggest priority will be staying on our mission, building on the phenomenal progress we made this year. Keep executing and keep hitting milestones. Demand will be huge, and meeting it will be challenging, but we must stay the course and not get distracted by the “noise”. All fundamentals are moving in the right direction. Let’s keep driving.
In closing, I want to extend my heartfelt thanks to everyone who has been part of this truly remarkable year. Wishing you all happy holidays, time to recharge, and a fantastic start to 2026.
– James Regan, CEO –
