Quantum of Alice (and Bob)

Théau Peronnin is explaining why he was so keen for his Paris-based quantum computing startup Alice&Bob not to contain the word ‘quantum’ in it.

Over the years, reasons the physicist and CEO, the word has been hijacked by fiction to come to mean the magical and the fantastical where reality defies the laws of mechanical physics.

But what Peronnin and his 40-strong team are building – and what they’ve just secured a Series A funding round of €27m for, is a marketable quantum computer with real world applications.

“There’s nothing magical about it – quantum physics is the most accurate representation of nature, without any magic in it and building a quantum computer is a well-defined challenge that requires an eloquent solution,” he says.

Hence Alice&Bob – a reference to two characters used in physics and cryptography theory (instead of A & B) – and a name which is proudly more physics textbook than quantum fiction.

And as the market moves from fantasy to theory to reality, there is real money pouring into quantum computing. A report by Global Markets Insights estimates that the market is set to surpass $5bn by 2028.

International Data Corporation meanwhile projects global customer spend for quantum computing will grow from $412 million in 2020 to $8.6 billion in 2027- a compound annual growth rate of 50%.

Little wonder then, that global technology giants – including Google, Amazon and IBM – as well as international academia and laboratories around the world  – are all racing to build the first commercial quantum computer.

In the latter category – a spin out from a federation of French research labs – is Alice&Bob, founded by Peronnin and fellow quantum physicist Raphaël Lescanne.

Quantum physics is the most accurate representation of nature, without any magic in it

As a start-up, explains Peronnin, it was important to take a “lean and singular and focused vision” by adopting a hardware-efficient approach to quantum computing.

“We’ve picked what we think is the main battle in this race, which is the quantum hardware. We’re building the chip devices that will be at the core of quantum computers,” he adds.

The simple aim is to build a quantum computer that does not make errors. One of the biggest challenges in building a quantum computer is the errors affecting qubits (a qubit, or ‘quantum bit,’ being the quantum version of the classic binary bit).

Companies like IBM are attempting to solve this problem by working on processors with more and more qubits.

Alice&Bob, however, have taken a qualitive approach by designing a superconducting ‘cat qubit’ circuit, which, it claims, is able to reduce the faults in processing thus reducing the number of processors needed to overcome these faults.

Recent work with one research lab, led by the experimental physicist Zaki Leghtas, has shown that these cat qubits resist “bit flips” (that’s the name for one of the two types of faults which appear in quantum computing) for up to several minutes.

While this might not seem long, the company claims that it’s a near “100,000 times improvement on earlier quantum efforts”.

With this recent discovery, Alice&Bob claims it has created “a shortcut to delivering a scalable universal quantum computer”-  by reducing the number of qubits required.

News of this breakthrough coincided with a successful $30m fundraising round, with investors including the French VCs Elaia, Bpifrance and Breega as well the disruptive tech fund manager Supernova Invest.

According to Peronnin, the plan now is to use the cat bit circuit to reduce the other known error in quantum computing: ‘phase flips.’ He estimates that this will take another year – meaning that the first fault tolerant quantum computer could be in circulation by 2023.

If it is a success the firm will deliver Quantum-as-a-service (QaaS) to allow businesses to utilise quantum computing resources remotely.

“No one so far managed to build this kind of quantum computer that manages to do things that a classical computer can do because of these errors. And so, in our case, we’re really focusing on solving the technical challenges,” he says.

Quantum’s got talent

Peronnin adds that one of Alice &Bob’s assets is its ability to attract talented and accomplished physicists and engineers to solve these challenges. To this end, the firm plans to hire a further 30 employees over the next two years, mostly with advanced physics skills, as well as developers and engineers supporting the commercialisation of its technology.

“The limiting factor in quantum computing at the moment is not funding – although it is quite capital intensive – it’s the access to scientists and to knowledgeable people that know how to build those extraordinary machines,” he says.

But even with scores of extra talent – does Peronnin really think he can compete with the might of Google and IBM? While it’s not about money yet, he admits that one day it will be and they will need to scale up. However, in terms of size, he adds that the start-up is not that far behind.

“If you look at Google’s quantum hardware team, it’s between 50 and 100 people from estimates. IBM is around 100 to 200 people. Now we are 40, we will be we aim to double within the next two years. So while we are not the same ballpark yet, but they are not lightyears ahead in terms of size,” he says.

While half Alice&Bob’s staff hold PHDs, you won’t see many sales and marketing people among its numbers. Peronnin appears so confident that the use cases for these machines are “ubiquitous,” that he says he’d rather focus on creating value to firm’s shareholders through improving the technology.

In terms of actual deployments, Peronnin can only speak in broad terms, but he sees three main areas where quantum computing will be used to solve problems that classical computers can’t. The first area is optimisation – optimising a financial portfolio or designing an optimal shape of aeroplane blade.

“Quantum computers can be used to speed up optimisation dramatically and find the ultimate solution, for example, for logistics or complex networks as well,” he explains.

The second area is in addressing engineering problems – or linear algebra – with quantum computers able to solve fluid dynamics equations in designing aeroplanes or to train neural networks for artificial intelligence.

“Most modern engineering could be exponentially speeded up and it has applications in sectors such as energy, engineering and data science,” Peronnin says.

However, what excites Peronnin the most is quantum computers’ ability to simulate molecules to decode secrets of genomes, drugs and proteins.

“Just to take a very practical example, if you want to simulate a molecule like penicillin – that’s a simple molecule – but it would take a classical computer, something like 10 to 80 free bits of memory, so more bits of memory than you have atoms in the universe.

“So, it’s not just that our current computers are not fast enough, they never will be. Whereas with a quantum computer, it will take something like 254 quantum bits, if I recall correctly, so while we’re not there yet, it’s all within reach.”