CEO Interview: Building a quantum computing leader in Europe

Interviews | January 11, 2022

By Nick Flaherty

Artificial Intelligence MPUs/MCUs Optoelectronics Materials & processes

French quantum computing technology developer Pasqal has merged with Qu&Co, a quantum algorithm and software developer in the Netherlands. The move marks the continuing consolidation in quantum computing and is a key step in building a leading European quantum computing company, CEO and co-founder Georges-Olivier Reymond tells eeNews Europe.

Continuing as Pasqal and headquartered in Paris, executives from the two firms say the combined company will show a 1,000-qubit quantum solution in 2023, a similar timescale to the announced roadmaps of the most advanced quantum platforms.

The company plans to have commercial 500 qubit machines by then for Qu&Co customers and partners, which include Johnson & Johnson, LG, Airbus and BMW Group.

Reymond will remain CEO of the combined company while Benno Broer, CEO of Qu&Co, will take on the role of Chief Commercial Officer. Broer also serves as vice president of the European Quantum Industry Consortium (QuiC). The merged company will have operations in seven countries and Qu&Co’s pan-European activities will continue under the names Pasqal Netherlands, Pasqal Germany, Pasqal UK and Pasqal Spain.

The key to the deal is integrating Qu&Co’s quantum algorithms with Pasqal’s neutral atom quantum hardware.

“Technically speaking its hardware buying software but its important to present it as a merger to create a European leader,” said Reymond.

This follows the merger of Honeywell’s quantum computing hardware operation with the algorithm development at Cambridge Quantum Computers to create Quantiniuum.

“The main difference between their deal and our is that Pasqal is a startup rather than a large corporate,” said Reymond. “From the beginning we had this idea of merging hardware and software, as the first applications will be hardware specific and to do this in an efficient way you need to know the hardware intimately. We have been doing that since the beginning with a strong software team and this merger is a way of speeding this up to broaden our application range and our team. We have 60 people and for a hardware company we will have over 25 employees working on software.”

The technology uses neutral atoms at room temperature manipulated with lasers rather than super-cooled ions or superconducting qubits. This allows the atoms to be moved around and re-configured on a cycle by cycle basis, which is particularly suitable for graph-based machine learning algorithms. The company is working with Muqans, a fellow spinout of Institut d’Optique d’Aquitaine in France, on the laser technology.

“We are leveraging neutral atom technology with single atoms as qubits and we do not need to manufacture them. They are operating at room temperature without complex cryogenic equipment and we have full control over the geometry to create quantum registers by positioning qbits in space,” said Reymond.

“So you can use the atoms to make a graph for instance and the shape of the graph changes from one application to another. Doing that, we have been able to build a device that can process a complete graph is five quantum operations rather than hundreds of thousands of operations.”

“We are using light to control the qubit and by scaling the devices with holographic techniques it is very flexible. Our roadmap is to reach 1000 qubits by 2023 and we currently have a set up with 500 atoms in the lab but we do not yet have the quantum control which will be ready by the summer,” he said.

“For a quantum machine learning algorithm we only need global addressing for the qubits as we address all of them in a single shot, that’s why its so efficient and that is very hardware specific. We prepare the quantum register, implement it and readout, then start again. It’s operating at a couple of hertz but we do more processing so its faster overall.”

“For gates between the quits you need individual addressing and we are using light and its definitely scalable and it has been demonstrated in the lab and the nice thing is the physics is easy to understand how it works. All the know how is how to put it together for the hologram and controlling the atoms and being able to move them and rearrange – we use off the shelf components.”

However this needs specialist algorithmic expertise. “So merging hardware and software is a way of speeding up the quantum advantage,” he said. “On top of that we are developing tools that are useful. “We have the APIs to be compatible with the Cirq Google platform, but we also have our own low level open source tool called Pulser, which is kind of assembly language,” he said. “

“On top of Pulser you can build libraries to solve mathematical problems such as finding properties on graphics and solutions to partial equations, then the next layer is the applications for automotive etc and then there are the end users, who can have access to the application layer, or to the controller at the Pulser layer,” he said.

The initial aim is to have a cloud-based quantum computing system, but large corporate customers are also looking for their own on-premises systems so hat they don’t have to send sensitive data to the cloud.

“In the short term we are scaling to the cloud for the end users to test the systems and evaluate the investment required and as we scale the technology and the business the first users will be ready to buy the hardware for their own use as they will not want to send their data to the cloud,” he said.

“We are already ready to sell devices, the first customers are the High Performance Computing (HPC) centres and they need to start exploring these new tools – we will deliver the first systems by 2023 with 100 to 200 qubits, the 1000 qubit will still be a prototype.”

Building a major European quantum computing company also takes cash. The company has raised over €25m ($30m) is backed by Quantonation with Runa Capital, Daphni and Eni Next, as well as the European Commission’s EIC investment fund, which includes an equity stake.

“We need strategic investors to validate the technology, and VCs to drive the money. We need European investors as it’s a strategic industry but the big money is in the US and we already have two of them.”

He is wary of investors demanding the sale of a company to get the return on investment. “This kind of strategic move prevented the rise of a European leader, as soon as a technology is ready they have sold on and I want to be a European giant able to compete with Google Facebook and Amazon and public investment is a way to achieve that,” he said.

But there is a long way to go to create a European leader in the industry, and there is a global race for funding. Psiquantum went to the US to raise $650m to commercialise its photonic quantum computing technology originally developed in Europe.

“Leveraging the synergies of the two companies will further strengthen our full-stack, neutral-atom solution in achieving near-term commercial benefit,” he said “Qu&Co’s technology and exceptional talent base has proven itself across several industries and will allow us to compete with any quantum computing company in the world. The technology is moving fast, the business is moving fast and the investors are moving fast,” said Reymond. “I think it is doable in Europe but you have to be careful. We have strong European investors and after the next round [of funding] we will still be a European company.”

www.pasqal.io

Related quantum computing articles

Other articles on eeNews Europe