We hear a lot about quantum computers – sometimes too much – but not as much about quantum networking which will also be a critical component in making widespread use of quantum information technology a reality. There are many important use cases including, for example, scaling up of quantum computers by linking them, providing quantum-secured secure communications, and networking together quantum sensors. There’s even a DoE-spearheaded Quantum Internet development project and most big cloud providers are also exploring the idea.

Getting from these early development days to a full-blown quantum internet will take time, but perhaps not as long as we think. Roughly two weeks ago, the first commercial quantum network – the EPB Quantum Network – debuted in Chattanooga Tennessee. Granted, it’s currently a playground for development. Still, it has everything needed for developers to sling photonic qubits around the network and start testing new products, and even sell access to their offerings.

(What is a quantum network? Good question. Broadly, it is the ability to send entangled quantum bits (qubits) over a network. Quantum networking is fast, secure, and promises to enable many applications. Included at the end of this article is a brief description by the Department of Energy. Also noteworthy, there are a handful of government/academic quantum network efforts.)

Developed by Chattanooga’s municipal power and communications company, EPB, with partners Qubitekk and Aliro, the new EPB Quantum Network is a quantum-as-a-service offering that provides quantum technologists with 216 managed, dedicated dark fibers and capacity for 10 quantum interconnected nodes across downtown Chattanooga. Designed by Qubitekk, the network architecture is configurable for customers and intended for use in validating quantum product performance, testing new quantum technologies, running quantum security applications and other uses.

“This is the nation’s first commercial quantum network and it really is helping to define what the infrastructure is going to look like, for quantum devices, you know, 5, 10, 15 years in the future,” said Duncan Earl, Qubitekk president and cofounder. “So this is definitely a very early service being offered in the in the quantum space. What we’re trying to do, on the technical side, is provide a network that users have really good access to, that they can subscribe to whether they’re a private company, a university or national lab, and it gives them access to technology that today is hard to get your hands on because it’s either expensive or just rare.”

“If you read any of the quantum papers, you hear about entangled photons, and Bell state measurements, and at the end of the day, there’s equipment that powers all of this. A lot of it is commercially available today. This network starts by having this ‘common-use’ equipment located in quantum data centers that allows users to leverage that equipment in addition to whatever equipment they’re bringing to integrate with the network. It’s a software defined network, so those resources can be used in many ways for many applications. It allows them to test their products or their solutions in a much faster way,” he said.

Earl and J. Ed Marston, of EPB, briefed HPCwire on new network last week. No early customers were named but Marston said, “At this point the users have primarily been the integrated technologies folks that Duncan has been describing. But we have been in conversations with a range of potential users from Fortune 500 companies, to major research institutions, and we think we’re very close to having some initial customers on-boarded and begin using the network. We’ve actually had more than 30 grant applications [in which] we were written in as the environment.”

Not surprisingly, EPB has high hopes for its new quantum network, including as a way to attract a broad range of quantum ecosystem companies to locate in the Chattanooga area. Access packages start at $10,000 a year. Presented here is a portion of HPCwire’s discussion with Earl and Marston.

HPCwire: Let’s start with some use cases. Can you give us an example of an early use case?

Earl: Let’s say you’re a small company building an offering that allows you to have quantum key distribution or secure communications. The way this used to be done is you would have to go to an organization (customer) and you’d have to sell them everything – so the equipment that allows you to do the quantum secure communications, you’d have to sell them on getting dedicated fiber in the ground to allow two endpoints to be connected through fiber, and you’d have to really sell this idea of integrating all of that with their larger communication system, and probably some development along the way. That is a very lengthy process. If you think about how you would connect banks together, it’s a huge infrastructure project, to put in this dedicated network to support that.

Those same companies can come to the EPB quantum network and instead of worrying about the fiber connections, and how you balance the quantum states as they go over optical fibers, they really just need to bring sort of 10% of the solution, connect that to this network through something we call the quantum network interface console, or the QNIC. It’s like a NIC card in your computer. That will allow them to connect to the rest of that network and leverage a lot of the resources that they already have for security applications. It reduces the cost of their equipment, it makes it much easier and much faster for them to deploy.

HPCwire: One of the complications about quantum systems today is there are so many different qubit modalities. Transducing from one modality to another, in this case into a photonic qubit isn’t trivial. Is that something that EPB provides? Do you have a way to interface with different kinds of modalities?

Earl: That’s a good question. In the early internet, a bit was a bit, so it wasn’t as complicated. Now, we’ve got all these different types of qubits. With the network that EPB has now, the primary qubit is a photonic qubit that interfaces with all of the nodes. The process where you go from a photonic qubit into a different modality of qubit – called the transduction process, which others are working on – we don’t solve that problem today. But it is definitely one of the use cases for developers using this network…to develop that transduction.

HPCwire: So, to get into the network, I have to have a photonic qubit to start with.

Earl: Fortunately, the network itself actually does generate photonic qubits. If you don’t have one (photonic qubit), it’s okay [because] that’s one of the common-use resources that is baked into the network. You can use software and say, “Okay, I don’t have my own source, but use whatever’s available on the network, generate this many qubits for me, and then I’m going to use them throughout the system.” If you have a quantum computer, and you want to use that photonic qubit, to work with say a matter-based qubit, the transduction between those two is still an active area of research. Actually, this network is a great place to do that research. But there are not yet commercial products to allow [different modalities] to talk to each other.

HPCwire: What kinds of companies are you’re thinking will do this kind of exploratory R&D work on the network? Must they be located in the Chattanooga, downtown area?

Marston: Duncan is more technical than I am about this, but I can take this question. The way the network is set up right now, and it is a fully quantum network, users do have to plug physically into this network. You can’t get to it over a classic network. We have a number of user nodes located in the downtown area. We also have some available spaces that people can lease on a longer-term basis. But the idea is that they would subscribe to the network and would basically have a set of hours where they could use it. They would need to be physically present and plug their equipment into it to use it.

HPCwire: I’m assuming that early on, most of this stuff will use the network as a round trip kinds of experiment – sending something out and get it back and see if in fact it did what they wanted to do. Is that wrong?

Marston: I think that’s exactly right. But one of the opportunities that’s available is a lot of this work is being done in isolation. Right now, it’s either in a corporate lab or university research facility or something. This network is designed so that all of the users retain their IP, they don’t have to worry about losing their IP when they use the network. That allows for collaboration. One of the great use cases is that they can bring their equipment in, show that it’s interoperable with another user’s equipment that’s installed on the network. It begins to create a point of convergence so that people can show that their equipment will work together.

We see this as an opportunity for folks who are working in isolation to come together. That collaboration can accelerate their ability to commercialize not only their [own offerings], but also the other folks’ that are working in the same space.

HPCwire: Given how young the quantum information industry is, are you hoping the network can attract others to create a quantum development hub in the Chattanooga area? Munich, for example, has what’s called the Munich Quantum Valley, sort of a nod to Silicon Valley.

Marston: That’s exactly what EPBs driving to do. We’re a local municipal electricity and connectivity provider. We’re a pioneer in the connectivity arena, and we really conceived this, with Duncan and Qubitekk, as a cornerstone for a new economic development effort. We see this as an opportunity to invite these companies to come in use the network. We are simultaneously working on building out a full-blown quantum ecology.

We did a large program last spring, celebrating World Quantum Day; we engaged over 180 teachers, reaching about 8000 students to raise their awareness about quantum. Duncan went into our local community college and one of our high schools and talked about the opportunity, just getting students interested in pursuing these fields. We definitely see this as a job creation opportunity and are connecting the dots with workforce preparation, education, and we’ve also connected with our local business accelerator, which is specializing in supporting quantum startups.

HPCwire: What are you thinking would be some of the early tests use cases?

Earl: I’m not going to dodge the question. But I want to mention there that Apple released their augmented reality glasses earlier this year. I don’t know if you saw that. But they’re very expensive. They didn’t really have any apps ready for it, because they were really releasing it for developers. [It’s] the developers who are going to build the use cases and aps. I think we have a similar situation here with the EPB quantum network. It’s really targeted not only at developers accelerating their product development, but also being able to float and trial early solutions very quickly. The collaboration part is really a big piece of that. There’s even some really neat spinoff ideas where the network can become a bit of a marketplace. So that early technologies can be tied together for a price, each user node to charge other user nodes to integrate their technologies. So it’s really a developer’s playground in some respects.

Having said that, there are clear applications that we know are coming. The security application where you use a quantum bits being shared between different devices to get this very secure communications. That is very strong in the rest of the world. We see that being embraced in Europe and in Asia, but it’s been a little slow to catch on here in the US. That’s partly because of the infrastructure challenges of doing that.

We also see some applications in time synchronization and other applications that play into nanosecond time synchronization. For example, themselves. Network to packet transmission where you know, exactly when it’s going to arrive, and it arrives at exactly that time, so that you don’t have anyone rerouting your traffic or maybe affecting a time sensitive process. In addition, as you mentioned, the really big sort of Holy Grail of these networks is tying together quantum devices, eventually, quantum computers. The early applications will really seek to try and advance that technology so that one day we can tie computers together. And as you probably know, on the quantum computing side, everybody’s, announcing bigger computers. If they had 32 qubits last year, and next year, they’re going to have 64. But the first guy that shows that we can take a 32-qubit processor, and over a network, tie it to another 32-bit processor, that’s a very, very scalable solution. That’s going to be a very attractive use case.

HPCwire: Does EPB do any quantum technology research? I’m thinking, for example, about quantum memory, which will be useful in repeaters.

Earl: There’s a very well defined roadmap we have for getting to the point where you can have distributed quantum computing. And you’re hitting some of the key product stuff and the early milestones. Quantum memory is one of the first key milestones. There’s already a number of companies that have very close to commercial solutions on the quantum memory piece. We think that this network will be the right way to test those out. Once you have a quantum memory, you have the ability to synchronize your photonic qubits that are going around the network. And so sometimes they call that on-demand qubit delivery, and that’s the second milestone – being able to integrate very narrow linewidth qubit sources with these quantum memories so we can deal with the synchronization. I could go hit the other ones too.

HPCwire: What does Qubitekk supply to EPB?

Earl: The network is actually made up of multiple vendors, every piece of equipment on the EPB quantum network is a piece of commercial technology, it has a vendor behind it. Qubitekk probably makes about 70% of that equipment, including the entangled photon sources, a lot of the measurement and preparation of those qubits as well. But we’re joined by other companies, DiCon Fiberoptics, which has been very strong, and in other markets, telecommunication markets, for example, they make a quantum-friendly or quantum-compatible, all optical fiber optic switch. Another company called Quantum Opus makes these very sensitive superconducting nanowire detectors. Qubitekk’s role is also as a system integrator.

HPCwire: The focus sounds very commercial.

Marston: Absolutely, the focus is commercialization. But we’re very well aware of where those technologies will originate, from national researchers to university researchers, certainly to entrepreneurs and corporate.

HPCwire: As I’m sure you know, Oak Ridge National Lab hosts lots of quantum research. And its nearby, relatively speaking.

Marston:  We’ve had a partnership with Oak Ridge which dates back to soon after 2010 when we launched the fiber optics, The Department of Energy named us as a living laboratory for smart grid technology, and we’ve been doing a range of projects with Oak Ridge and other national research institutions ever since. In fact, how we got to know Duncan and his company, was jointly working on a quantum cryptographic technology. The US Department of Energy was requesting research for securing the national power grid. And together with Oak Ridge and Qubitekk and Los Alamos we actually got a R&D 100 award for that early development work. The experience then led to launching EPB quantum network.

HPCwire: Thank you for your time.

Related HPCwire coverage: Catch the Flying Qubit  – AWS Center for Quantum Networking, https://www.hpcwire.com/2023/07/11/catch-the-flying-qubit-aws-center-for-quantum-networking/

What is Quantum Networking?

(Excerpt from DoE)

Because there are new scientific domains to explore. Quantum physics governs the domain of the very small. It allows us to understand – and use to our advantage – uniquely quantum phenomena for which there is no classical counterpart. We can use the principles of quantum physics to design sensors that make more precise measurements, computers that simulate more complex physical processes, and communication networks that securely interconnect these devices and create new opportunities for scientific discovery.

Quantum networks use the quantum properties of photons to encode information. For instance, photons polarized in one direction (for example, in the direction that would allow them to pass through polarized sunglasses) are associated with the value; one, photons polarized in the opposite direction (so they don’t pass through the sunglasses) are associated with the value zero. Researchers are developing quantum communication protocols to formalize these associations, allowing the quantum state of photons to carry information from sender to receiver through a quantum network.

Quantum networks use uniquely quantum phenomena, like superposition, no-cloning, and entanglement that are not available to classical networks. Before the photon is measured, it exists in a superposition of all its possible quantum states, each with a corresponding probability. Measurement selects one among these states. In fact, the photon’s quantum state cannot be measured without causing a disturbance that betrays the attempt. Nor can an arbitrary, unknown quantum state be copied – no cloning allowed. A properly designed and operated quantum network derives inherent security from this behavior.

But if the photon cannot be copied, how can the communication be amplified to reach distant recipients? This is where the quantum phenomenon of entanglement enters the picture. The quantum state of each entangled photon is correlated with that of its entangled partners, regardless of their distance apart. Quantum network repeaters are being developed that use entanglement to extend the range of quantum networks.

Will the emerging quantum internet make today’s classical internet obsolete? Not at all. The strengths of quantum networks are complementary to those of classical networks. We will reap the greatest benefit in the long run by incorporating both classical and quantum networks in an internet with capabilities that exceed what is possible with either technology on its own.