For many scientists and engineers, Alice and Bob are two fictional characters commonly used when creating the scenario for a research experiment that involves communicating between a point A (Alice) and a point B (Bob).
Paris-based quantum startup Alice & Bob, which designs and builds quantum computers, borrowed the names for the sake of the private joke, says CEO Théau Peronnin — but also to intentionally avoid any obvious reference to quantum technologies.
“The word ‘quantum’ is largely misused in popular culture to say ‘magical’,” says Peronnin. “But there is nothing magical about what we are doing.”
Advertisement
He might be modest, but Alice & Bob is a company that’s making a mark on Europe’s quantum industry, recently landing deals with the likes of the French government and US chip giant Nvidia to accelerate the race to build useful computer systems with the technology. To date, it’s raised €30m from French VCs Elaia, Breega and Supernova Invest, as well as the European Innovation Council (EIC) fund and French public bank Bpifrance.
Scientists predict that quantum computers will one day unlock significant computational power, surpassing that of today’s conventional supercomputers and bringing huge value to businesses.
Startups and big tech companies around the world are racing to build a fully-fledged quantum device, using different techniques based on various scientific theories — and all hoping that their own approach will show an advantage of scale or efficiency over others.
Although Alice & Bob is only four years old, Peronnin is confident that the company can outpace some of the world’s biggest technology companies — and find the capital necessary to build multimillion-dollar experimental devices.
That is no mean feat for a French startup, he tells Sifted, with quantum considered a “critical technology” by many European governments, limiting the company’s exit options and making fundraising all the harder.
200x better qubits?
Different companies are working on different ways to build a quantum computer, and more specifically the “qubits” that power them. These are the quantum equivalent of “bits” — a mechanical system that’s used to store and encode computational information.
Alice & Bob is building its systems with ‘superconducting qubits’, the same broad approach being used by tech behemoths IBM and Google.
Both of those companies have been actively developing quantum computers — and pouring billions into the technology — for the past decade. Alice & Bob launched in 2020 with a €3m seed round — but Peronnin sees this as a strength.
“We are like a second generation of the [superconducting] approach,” he says. “We got started with a redesigned qubit… that enabled us to correct natively one of the two errors that qubits typically suffer from.”
Qubit errors, which are caused by the components’ high instability, are one of the main challenges holding back the development of quantum computers. Currently, around 1k qubits are necessary to create one fully ‘error-corrected qubit’, or ‘logical qubit’ — meaning that scaling up quantum computers is still a significant engineering challenge.
Advertisement
Alice & Bob says that it has designed a new type of qubit, called a ‘cat qubit’, which has built-in correction for a big chunk of these errors. The company says that it has theoretically demonstrated that only five of its cat qubits could eventually be necessary to build one logical qubit.
If it can prove that in the real world, a 200x more efficient qubit correction rate would be significant.
But the startup isn’t alone in the business of building cat qubits: at the end of 2020, Amazon’s cloud business AWS started focusing on the same technology to make quantum computers.
Alice & Bob is therefore directly competing against AWS on hardware development — and the battle is on for talent. “They pay about 2.5 times better than us,” says Peronnin, “but we haven’t had any defections yet, and we’ve even managed to attract people that had offers from both sides.”
His theories as to why his company is able to compete on such uneven financial terms include the idea that some are tempted by the “David and Goliath” environment of working at a plucky startup, and its central Paris HQ doesn’t make for a bad quality of life either.
Generating revenue
In its inner city lab, Alice & Bob currently hosts four quantum computers, ranging from one to 16 cat qubits.
None of these devices has yet achieved a logical qubit. Peronnin says that the company’s objective is to create one logical qubit in the next two years, and 100 by 2028, enabled by 1,500 cat qubits.
This is when the company expects to start unlocking initial use cases, mainly for research purposes. Industrial applications, says Peronnin, will come at a later stage, requiring several hundreds of logical qubits.
For now, however, the technology is in its early stages — and has very little commercial value. Peronnin says that last year Alice & Bob generated no revenue, and that the company is only aiming to be profitable “by the end of the decade”, to coincide with the emergence of first use cases.
This is a different approach to some of the startup’s competitors, like Paris-based Pasqal, which are already bringing in significant revenue by selling their hardware and securing industrial partnerships to support businesses’ quantum strategies, even while the technology is still immature.
“We’re holding our breath for longer,” says Peronnin.
“We’ve decided to focus directly on where we are certain that the impact will be massive — fully-fledged, error-corrected machines — without getting distracted by intermediate devices that have not been mathematically proven to serve a use case.”
Funding for quantum
Alongside its VC funding, Alice & Bob has also recently received a grant from the French defence ministry to participate in a competition programme against four other startups — including Pasqal and Quandela — to build a quantum computer with 128 logical qubits by 2028. As part of the programme, Alice & Bob will pocket about €20m over the next two years, according to Peronnin.
But the startup will need more cash to fund its roadmap, which includes building new devices and opening a new production site. Two sources tell Sifted that the company has started raising a €100m Series B.
“We don’t have a cash problem thanks to the grant we received, but we will eventually have to raise a Series B,” says Peronnin. “And the order of magnitude, Pasqal has set the tone, is €100m.”
It’s not the kind of money that comes easy in Europe.
“Europe is very badly equipped to do this,” says Peronnin. “If I’m looking for a lead investor that can write a €50m cheque… it means I’m looking for an investor with a €1bn fund that can back deeptechs.”
“In Europe, there aren’t many of these.”
The exit challenge
Peronnin says that the startup is likely to look for investors in North America; and last year, Alice & Bob opened a subsidiary in Boston to facilitate the company’s expansion.
But a European quantum startup can be a hard sell for a foreign investor. As a company building a technology in a sector that is considered critical to the bloc’s sovereignty, it is likely that Alice & Bob will be heavily scrutinised in the event of an exit, for example if it receives an offer for M&A outside of Europe.
This significantly limits the startup’s exit options that, otherwise, might have looked bright.
“We’re on the same platform as tech giants, but with a better design and a good pool of talent,” says Peronnin. “We are an alluring target for acquisition, so it is a big door that is closed from the start.”
“It’s normal, but it makes my job a bit harder.”
Peronnin says that the rules of the game were clear from the start. “By default, it will be an IPO,” he says.
This is not on the cards just yet. “Publicly listing in quantum computing today would be very difficult,” says Peronnin. “There is no revenue and no device delivering fully on its promises.”
Instead, the founder anticipates an IPO could happen towards the end of the decade, once the company has demonstrated the usefulness of its technology and seen the start of commercial adoption.
“We’ll need to have done two or three device sales for a few dozen million euros each,” says Peronnin, ‘to have a nice story to tell at that point in time.”
