29-05-2025
Quebec startup shows progress toward practical quantum computing
Julien Camirand Lemyre wants to correct the errors in his way.
To be clear, this is not a quest for personal improvement. It's a technical challenge and, for the nascent quantum computing industry, an extremely important one.
Mr. Lemyre is a PhD physicist and chief executive officer of Nord Quantique, a startup based in Sherbrooke, Que. Since 2020, he has set his company's sights on overcoming a key obstacle that stands in the path of commercial quantum computing: the technology's propensity for making mistakes.
On Thursday, Nord Quantique announced it had taken an important step on its path toward surmounting that barrier. The company has successfully used one of its own quantum devices to encode a form of error detection for the first time.
Bigger players, including Google, Microsoft and Amazon, are working on the same problem as they seek to advance their own quantum systems. What's different about Nord Quantique is that the hardware doing the checking is the same hardware doing the calculating.
The experimental result suggests that larger, commercially relevant quantum computers can be constructed from similar components. If so, those computers might occupy only a modest amount of space – something like a standard data centre rather than a football-field size complex that some fear will be required to get other types of quantum systems to run reliably.
'We think there are better ways to quantum error correction,' Mr. Lemyre said. 'This ties in with our philosophy of really working on something that we think is worth scaling up.'
The company's announcement, together with an accompanying scientific paper, is the latest step in what has become a industry-wide push to tackle error correction, also called fault tolerance.
Fault tolerant quantum computers have yet become a practical reality, but they are an attractive business proposition because they are expected to one day perform various kinds of calculations that are out of reach of conventional digital systems. Potential applications range from data security, to drug discovery to forecasting, among other areas. Yet the same properties that make quantum computers powerful also make it easy for them to fail.
Ordinary computers use bits – the electronic components that represent the 1s and 0s of a digital operation. In a quantum system, the bits are replaced with qubits, which are more versatile and more finicky. Thanks to the slippery rules of quantum physics a qubit needn't be a one or a zero, but can be a bit of both. But this ambiguous state of being, so essential for quantum computation, is easily disrupted by outside influences such as vibration or heat.
Microsoft creates chip it says shows quantum computers are 'years, not decades' away
The standard way of dealing with this is to dedicate other qubits to keep tabs on the first one. But this gets complicated and costly. For every qubit required to perform a calculation, more than 1,000 may be required for error correction. Imagine a Hollywood celebrity with an entourage that would fill an entire hotel and you can see how the problem multiplies as more celebrities join the party.
Nord Quantique uses a different kind of qubit than many other systems, involving microwaves in a supercooled cavity. The microwaves consist of individual particles, or photons, that have different ways of bouncing around in the cavity called modes. What Nord Quantique has shown in its latest work is that these modes can be used for a type of error detecting code called Tesseract without the need for additional hardware.
Mr. Lemyre said there are ways in which the approach can be further improved, such as by adding more photons to the cavity. And the system would draw only a fraction of the energy needed by other approaches.
Yvonne Gao, an assistant professor at the National University of Singapore who is familiar with the company's work, said that the work represents good progress along one possible path toward a fault tolerant quantum computer. She said Nord Quantique has helped the field by adding to the diversity of approaches to error correction, while carving out its own niche.
'It's a very smart choice not going head on with the other people working on other flavours' of the problem, she said.
While some larger companies have made huge investments in quantum computing, it is unclear which approach is most likely to succeed. That means smaller startups with novel technologies to explore may ultimately be the ones who find the way forward.
Three Canadian companies vying for U.S. quantum computing funding as race to develop technology heats up
Daniel Gottesman, a theoretical physicist at the University of Maryland who played a part in developing the codes that Nord Quantique and others are using for error correction, said that it was surprising that no clear winner has yet emerged among the various approaches being tried.
One reason for this, he said, is that the difficulty in building and controlling such systems is challenging enough that even the best-resourced companies cannot zoom ahead, but instead must work methodically at improving error rates and increasing the number of qubits in their devices.
'That takes time and gives other people time to do that work as well,' he said.
Nord Quantique is not the only Canadian company in the error correction game. Last February, Photonic Inc. of Coquitlam, B.C., publicized its approach to the problem, which builds on an alternative strategy for tying qubits together known as QLDPC (quantum low-density parity check) codes.
This class of codes is well suited to Photonic's quantum computing hardware, in which qubits are based on the spins of carbon atoms that reside within silicon chips. Because the chips can be interconnected with light guided by fibre optics, the qubits do not need to be physically adjacent to one another to be linked. This means the task of error correction can be spread out, creating opportunities to harness groups of qubits in more efficient ways.
Housed in a non-descript industrial unit east of Vancouver, Photonic has grown its head count to 150 since coming out of stealth mode 18 months ago. The company is now preparing to expand into a larger building next door to facilitate its hardware development.
Together with Nord Quantique and Xanadu Quantum Technologies Inc. of Toronto, Photonic is one of the Canadian companies to be selected by the U.S. Defense Advanced Research Projects Agency (DARPA) to compete for support in developing quantum computing technologies.
Stephanie Simmons, who founded Photonic in 2016 and leads its technology development, said that despite the challenge, the reason for the increasing sense of excitement in the field is clear.
'Every time you commercialize a branch of physics it changes everything,' she said.
