Quantinuum – the newly-named company resulting from the merger of Honeywell’s quantum computing division and UK-based Cambridge Quantum – today launched Quantum Origin, a service to deliver “completely unpredictable cryptographic keys” based random numbers generated by a quantum computer. Quantinuum calls Quantum Origin the first commercial quantum cryptography product of the NISQ (noisy intermediate scale quantum) computer era.

“It is here today, it exists. It’s not two years, or five years, or 10 years, or 20 years in the future. It is a product that cannot be generated or created using a classical computer,” said Quantinuum CEO Ilyas Kahn during a ‘virtual’ roundtable held yesterday on the eve of the Q2B conference, which opens today at the Santa Clara Convention Center.

Duncan Jones, head of quantum cybersecurity at Quantinuum, said, “In a nutshell, Quantum Origin is a cloud-based platform that uses a quantum computer, specifically, the H1 series (trapped ion) quantum computer from Quantinuum, to generate completely unpredictable cryptographic keys. Our cryptographic keys are the strongest cryptographic keys that have ever been created and also [that] could ever be created.”

Kahn emphasized users of the service don’t need quantum hardware or expertise. At least to start, an API will be all that’s needed. It’s basically a black box to users.

Is this an early example of achieving quantum advantage – the moment when an application’s performance on a quantum computer is sufficiently better than on a classical computer that it’s worth switching?

Broadly, Quantum Origin is a quantum random number generator – often just called QRNGs. That may sound trivial, but it’s not, and there are many who think these QRNGs and the applications they can enhance represent the low-hanging fruit in quantum computing. Zeroing in on cyber cryptography as a first market makes sense as recent hacks such as the Colonial Pipeline have painfully pointed out. Truly random strings of zeros and ones are the ideal starting points in cryptography.

“The way we measure the quality of an encryption key,” explained Jones, “is simply by how random it is, how unpredictable that key is because cryptographic keys are pretty much just random strings of zeros and ones. You want to be in a situation where your attacker has absolutely no idea what your key is [and] cannot predict a single thing about it.”

Quantum computing, based on superposition and entanglement, can deliver such strings. “If you create a quantum state that is 50 percent zero and 50 percent one, when you measure it, you cannot, in any sense, predict what value you’re going to get. You’ll get either a zero or a one, but you have no chance of determining which one it’s going to be, and this remains true, even if you have a supercomputer,” said Jones. The Continuum solution here is actually a hybrid quantum-classical system with the quantum computer contributing the decisive advantage.

Computing even these relatively straightforward random number generating circuits on a quantum computer isn’t easy. The usual list of challenges associated with quantum computing, such as system and environmental noise, can disrupt the circuit. Quantinuum says it has developed the necessary quantum system control and circuits to reliably produce random numbers. Currently, it’s using its own H1 trapped ion quantum computers. There are, of course, many qubit technologies being developed (trapped ion, superconducting, optical, cold atoms, etc.).

Asked if trapped ion technology offered a particular advantage Kahn said, “It’s too early to tell? We’ve tested on a wide variety of computers. We’ve tested on the Oxford Quantum Circuits, a superconducting machine. The original beta test of this was in October of 2020, using an IBM computer in the ecosystem that they generate.”

“What we’re finding is that the trapped ion device, because of its fidelity, is giving us unheard of scores [that are] a reflection of what is called either the Mermin test or the Bell test. These are agreed [upon] measures defining the quality of the randomness. Now there are other things which come into play such as output amount, and speed, etc., but on balance at the moment, the best results have come from the trapped ion [device]. But that could change tomorrow,” he said, acknowledging these remain early days among competing qubit technologies.

Quantinuum reported the new service has already been used in a series of projects with launch partners. Two of its early partners – Axiom Space and Fujitsu – were present at the roundtable. Axiom Space, which is building a commercial space station, used Quantum Origin to conduct a test of post-quantum encrypted communication between the ISS and Earth — sending the message “Hello Quantum World” back to earth encrypted with post-quantum keys seeded from verifiable quantum randomness. Fujitsu integrated Quantum Origin into its software-defined wide area network (SDWAN) using quantum-enhanced keys alongside traditional algorithms.

Continuum is following industry practices with regard to pricing said Jones.

“When you go to a cloud service provider, and ask AWS or Google or somebody to generate you a cryptographic key, they will typically charge you $1 a month to have that key for you and hold it in a secure environment. This is a standard key generated using the means that we have now surpassed. Now a typical key will be alive for about five years or so, which is 60 months. So you’re generally spending $50-$60 on a key,” Jones said. “Quantum Origin, is [also] charging on a per key basis and we charge a fraction of that overall lifetime cost of a key for the added value of ensuring that key is completely unpredictable.”

There were many questions at the roundtable about security and technology export control. Kahn said Quantinuum was following various regional government requirements and that the customer on-boarding process was rigorous. Cybersecurity, and data security in general, is an area of urgent research and activity. The National Institute of Standards (NIST) in the U.S. and the National Cyber Security Centre (NCSC) in the U.K. have active programs.

NIST has been actively developing so-called post-quantum standardization (algorithms) intended to impede the ability of quantum computers to hack encrypted messages. In theory, fully fault-tolerant quantum computers will be able to decrypt any classically encoded keys. NIST efforts are currently based on classical systems and fault-tolerant quantum computers are not expected soon.

Kahn said, “Looking into a crystal ball, I think that there will be frenetic activity on post-quantum algorithms. Why do I say that? None of the existing NIST candidate algorithms have ever contemplated or are [they] being designed to use output from an actual quantum computer. They’re deterministically evolved and based on computational complexity assumptions, that a quantum computer cannot do X, Y, or Z. That boundary, if breached by an algorithm, makes it safe from a quantum computer. Once we provide verifiable quantum output, which is unpredictable, algorithms that we can’t even conceive of will emerge.”

Circling back to QRNGs generally, there are many application areas that can be improved by using truly random numbers. One quantum start-up, Harvard spinout Zapata, is building its early business model around just that idea of leveraging quantum-generated random numbers as input for various applications.

Jones noted, “Quality randomness is a critical component in many different systems, particularly in the world of simulations. Monte Carlo simulations, for example, need access to high quality randomness. So yes, I think they will be potentially many use cases for this beyond cryptography. Perhaps the better question might have been why focus then on cryptography first. I think [it’s] just the place where randomness is most critical now, but yes, I can imagine we may provide randomness for other use cases in the future.”

Kahn added, “There is a use case that we have previously highlighted and it’s in the very early stages of development. We’ve been providing CERN with output, effectively from Quantum Origin, because they’re interested in large-scale Monte Carlo simulations. Whenever you get into this sort of really huge amounts of data that are being analyzed, [and] use Monte Carlo simulation methods, it is posited that you may get a better result if the quality of the randomness used is truly unpredictable. I think it is early days yet. I’m not sure we could say with any definite consequence that this is going to be a use case, but the intuition we have is that [quality randomness] will be useful outside of cybersecurity.” Machine learning is another frequently mentioned area that may benefit.

Currently Quantinuum has a staff of roughly 400 people. In its earlier launch notice the company reported “Quantinuum will globally launch a quantum cybersecurity product in December 2021, and later in 2022, an enterprise software package that applies quantum computing to solve complex scientific problems in pharmaceuticals, materials science, specialty chemicals and agrochemicals. It will also announce major upgrades to the System Model H1 hardware technologies.” Presumably these near-term offerings will also leverage QRNGs.

Link to Quantinuum release: https://www.quantinuum.com/pressrelease/introducing-quantum-origin