Quantum computers usually mean massive machines, cryogenic temperatures, and isolated data centers. But what if quantum computing could run at room temperature, fit inside a server rack — or even a satellite?
In this episode of TechFirst, host John Koetsier sits down with Marcus Doherty, Chief Science Officer of Quantum Brilliance, to explore how diamond-based quantum computers work, and why they could unlock scalable, edge-deployed quantum systems.
Topics include:
- Why diamonds are uniquely suited for quantum computing
- How NV centers work at room temperature
- Quantum sensing vs. quantum computing
- Manufacturing challenges and timelines
- Quantum computing at the edge (satellites, vehicles, sensors)
- The future of hybrid classical-quantum systems
Marcus explains how nitrogen-vacancy (NV) centers in diamond act like atomic-scale qubits, enabling long coherence times without extreme cooling. We dive into quantum sensing, quantum machine learning, and why diamond fabrication — including the world’s first commercial quantum diamond foundry — could be the key to manufacturing quantum hardware at scale.
You’ll also hear how diamond quantum systems are already being deployed in data centers, how they could operate in vehicles and satellites, and what the realistic roadmap looks like for logical qubits and real-world impact over the next decade.
- Get deep insights on the TechFirst Substack
- Subscribe to the TechFirst YouTube channel
Check out our conversation here:
Quantum computing via diamonds: transcript
Note: this is a partially AI-generated transcript. It may not be 100% correct. Check the video for exact quotations.
John Koetsier:
Are diamonds going to power the next wave of quantum computing? Hello and welcome to Tech First. My name is John Koetsier. My guest today is a quantum computing scientist whose best friend might just be diamonds. Diamond-based quantum computers don’t need to be frozen to unearthly temperatures. They enable long coherence, they have compact hardware, and they can live right alongside classical hardware. But are they the future of tech?
We have the Chief Science Officer of Quantum Brilliance. He’s a university professor, he’s an Army Reserve officer—which is interesting, the first quantum scientist I’ve talked to with that on his résumé—and also an entrepreneur.
His name is Marcus Doherty. Welcome, Marcus. How are you doing?
Marcus Doherty:
Yeah, brilliant. Thanks, John. Thanks for having me.
John Koetsier:
Excellent. Brilliant. I don’t know if that’s Australianism or just Quantum Brilliance being brilliant, but hey, it works. It’s 6:00 a.m. for you. Crazy hours. You’re in Australia, I’m in Vancouver, Canada. Glad we could make something work.
Let’s start really general. Why diamonds? Why are you using diamonds for quantum computing?
Marcus Doherty:
Yeah, well, as you always say in your intro, John, they have this remarkable property that you can do quantum computing and quantum sensing entirely at room temperature and pressure.
The reason they can do that is you’re using a defect in diamond called the nitrogen-vacancy center. That’s where you take a carbon atom in the diamond, replace it with nitrogen, and then remove the neighboring carbon atom to create a vacancy. That creates a nitrogen-vacancy defect, which behaves exactly like an atom.
You can engineer these atomic-like objects inside diamond and place them where you want them. And because diamond is so hard, even at room temperature there isn’t enough thermal energy to create lots of vibrations and disturbances in the crystal. As a result, you don’t get much decoherence, because vibrations are often what create decoherence in quantum systems.
That’s the fundamental reason: the ability to engineer qubits exactly where you want them, and have them operate entirely at room temperature.
John Koetsier:
That’s super interesting. The first quantum computing center I went to was in Waterloo, Ontario. I remember going into that building and how much they emphasized the engineering—how isolated it was from the world, how they spent tens of millions of dollars just making it stable.
If there was an earthquake, a truck passing by, or even slight tremors, the building wouldn’t shake so the qubits wouldn’t decohere. Now you’ve got all of that inside a diamond. That’s crazy.
Marcus Doherty:
Yeah, absolutely. The diamond is essentially giving you everything that that engineering was doing artificially.
John Koetsier:
Nice. Now, you’re making your own diamonds, right? You established the world’s first commercial quantum diamond foundry just a month ago or so.
Marcus Doherty:
Yeah, that’s correct. Diamond isn’t yet at the same status as silicon in terms of maturity, accessibility, consistency, and quality.
You can’t yet buy diamond wafers the way you can silicon wafers and use them to create millions of devices per year through a standard semiconductor foundry. One of the key building blocks of the diamond quantum technology industry is making that possible—creating a foundry that can produce diamond at the right scale and the right quality.
Our quantum diamond foundry is a step in that direction, creating a commercial pipeline for quantum-grade diamond.
John Koetsier:
Can you create them defect-free—super pure, super high quality?
Marcus Doherty:
Yeah. That’s the baseline. Getting defect-free, ultra-high-quality diamond.
This means removing isotopes like carbon-13 so you’re left with pure carbon-12 diamond. On top of that baseline material, you then introduce functionality. You need to introduce the NV centers, which are the defects, and you do that through different methods.
To support those NV centers, you also need n-type and p-type diamond, just like in silicon transistors where you dope regions to create p–n junctions. Diamond needs the same thing.
John Koetsier:
It’s always a good backup plan—if quantum computing doesn’t work out, you’ve got diamonds.
Okay, you talked about some of the advantages of using diamonds. How does it actually work?
Marcus Doherty:
Diamond qubits can be used for sensing or for computing.
For sensing, we create lots of NV centers—millions of them. They’re all identical, but they’re far enough apart that they don’t interact with each other. That’s important.
When you do quantum sensing, you address all of them at once. You prepare them to be sensitive to something in the outside world, like a magnetic field. They’re exposed to that environment, their quantum state changes slightly, and then you read them out. You can detect extremely small changes caused by that external field.
You use light to initialize the NV centers into a qubit state, microwaves to prepare the sensitive state, and then light and electrical detection to read out the final state.
For quantum computing, the step you need to take is placing NV centers exactly where you want them, close enough to interact with each other. That allows you to do two-qubit operations and create entanglement, while still being able to address each one individually for single-qubit operations.
It’s very similar to laying out transistors on a classical processor.
So today, we can build quantum computers with a few qubits, and quantum sensors with millions of qubits. The roadmap is to refine our engineering so we can manufacture sensors at scale, then translate that capability into quantum computers with millions of qubits.
John Koetsier:
What does that roadmap look like? Years or decades?
Marcus Doherty:
We see it as five to ten years.
Right now, Quantum Brilliance has created quantum computers that can be packaged as desktop-sized systems and integrated into data centers. They’re only a few qubits, but they’re real.
On the sensing side, companies like Bosch, Lockheed Martin, Q-CTRL, and SandboxAQ have built diamond quantum sensors. These are field-deployable prototypes. For example, Q-CTRL demonstrated magnetic navigation using a diamond quantum sensor on a U.S. military flight across the Pacific, achieving around 100-meter accuracy.
The challenge now is moving from prototypes to manufacturable products.
Our roadmap is focused on creating integrated quantum chips—the equivalent of the silicon chip for classical electronics. That means building manufacturing lines for quantum sensors first, then using those same tools to enable large-scale quantum computing.
By 2030, we aim to produce around 100,000 quantum sensing units per year. By 2028, around 1,000 units per year. In 2031, we aim to demonstrate our first logical qubit. By 2033, the goal is to produce 64-logical-qubit devices at scale.
John Koetsier:
How many physical qubits do you need for one logical qubit?
Marcus Doherty:
Honestly, we don’t know precisely yet.
Using conventional estimates, with fidelities above 99% and two-qubit operations around 99.9%, it might be around 1,000 physical qubits per logical qubit. But the error correction methods for diamond quantum computers are closer to those used in neutral atom systems than superconductors, so the exact number is still being worked out.
John Koetsier:
What’s fascinating is that you’re delivering systems today that fit on a desk. You’re also one of the few companies talking about quantum at the edge. Where does that make sense?
Marcus Doherty:
Right now, our systems are desktop-sized—19-inch rack units. We’ve delivered them to Germany, the U.S., and Australia. In the U.S., we’ve delivered three systems to Oak Ridge National Laboratory.
Inside each unit, we combine the quantum processor with GPUs and CPUs—there’s an NVIDIA GPU and a conventional CPU in the same box. That allows researchers to explore hybrid systems and parallel clusters of quantum computers.
The future we see looks a lot like classical computing today: instead of one quantum computer in a data center, you’ll have thousands. And not just in data centers—also in vehicles, medical suites, satellites, and maybe even homes.
This complements mainframe-style quantum computers. The future will be a mix: our technology doing many small tasks at scale, while other platforms handle fewer, highly specialized tasks.
John Koetsier:
If you put one of these units in a vehicle or satellite, what is it actually doing?
Marcus Doherty:
One of the clearest near-term advantages is quantum machine learning.
Quantum machine learning can deliver similar or better accuracy with fewer training cycles and less data. It’s also more robust to noise, both in training and deployment. For example, in image classification tasks with fog or interference, quantum models can maintain accuracy better than classical ones.
Quantum machine learning works just like classical machine learning in one key way: you need thousands of quantum computers to train the model, and then either quantum computers at the edge or many in data centers to deploy it.
A concrete example is our work with Lockheed Martin on synthetic aperture radar. Satellites collect electromagnetic signals and reconstruct radar images of ships at sea. There’s limited training data and lots of background noise—perfect for quantum machine learning.
Doing that processing directly on the satellite, instead of sending data back to Earth, is incredibly powerful.
John Koetsier:
That makes sense. The data is right there, processed instantly.
How many processors could you fit in something like a satellite or vehicle?
Marcus Doherty:
We’re targeting a PCIe form factor—the same size as today’s GPUs. Inside that would be a single processor chip with 64 logical qubits by 2033.
For that satellite task, you’d have one of those onboard, and then thousands of them in a data center on the ground for training.
John Koetsier:
Are there disadvantages to using diamonds for quantum computing?
Marcus Doherty:
Building qubits is hard—no matter the technology.
The reason we talk about timelines like 2033 is that it takes time to mature the fabrication processes. It’s very similar to the silicon transistor era in the 1960s, when it took a decade to build reliable foundries and processes.
Diamond quantum computing is a second-wave technology. Other systems are exquisite but difficult to manufacture at scale. Our approach takes longer upfront, but once it’s there, it can scale to millions of devices.
And importantly, we don’t have to wait. Quantum sensors are already useful products, and they help build the manufacturing foundation for future quantum computers.
John Koetsier:
Super interesting. For years, I’ve talked to quantum companies building fridge-sized, room-scale systems. Now you’ve got quantum computing in a box that can go into servers, vehicles, and satellites—and you can buy a two-qubit system today.
Marcus, thank you so much for taking the time. I really appreciate it.
Marcus Doherty:
No problem, John. Real pleasure.