QuantrolOx uses machine learning to control qubits

Share
  • February 24, 2022

QuantrolOx, a new startup that was spun out of Oxford University last year, wants to use machine learning to control qubits inside of quantum computers. The company, which was co-founded by Oxford professor Andrew Briggs, tech entrepreneur Vishal Chatrath, the company’s chief scientist Natalia Ares and head of quantum technologies Dominic Lennon, today announced that it has raised a £1.4 million (or about $1.9 million) seed funding round led by Nielsen Ventures and Hoxton Ventures. Voima Ventures, Remus Capital, Dr. Hermann Hauser and Laurent Caraffa also invested in the round.

The company’s technology is technology-agnostic, and could be applied to all of the standard quantum computing technologies. The idea here is that instead of going through a slow manual tuning process, QuantrolOx’s system will be able to tune, stabilize and optimize qubits significantly faster. Current methods, QuantrolOx CEO Chatrath argues, aren’t scalable, especially as these machines continue to improve.

“I was talking to one U.S. investor. He said that we are like the picks and shovels of the quantum industry, in that we don’t have to wait to get revenues for a quantum computer to be useful,” Chatrath said. “As you get from five qubits to — hopefully — millions of qubits, you need our software every single day to be able to do the device characterization and tune the qubits.”

For the time being, though, the company’s focus is on solid-state qubits. In part that’s because those are systems the company has access to, including through a close partnership with a lab in Finland that the company wasn’t quite ready to disclose yet. As with all machine learning problems, QuantrolOx needs to gather enough data to build effective machine learning models.

As Chatrath also noted, we’re still in the very early stages of quantum computing, but if tools like QuantrolOx can help researchers speed up the process of testing their devices, that’s a boon for the entire industry. He noted that a lot of companies in the industry are already approaching the company to use its control software.

The company currently has seven full-time employees and plans to hire about 10 more people in the near future. But as Chatrath noted, he doesn’t expect that number to grow much more in the next two years. “We don’t need a huge team, because we are focusing on our specific niche,” he said. “We don’t want to full-stack. We don’t want to go higher in the stack — and we can’t go lower in the stack because that’s the hardware. So we are very much focused.”

Currently, QuantrolOx is focused on building more partnerships with the builders of quantum computers. These are rather deep partnerships because the team essentially needs access to the physical machines but also the source code that controls them so it can integrate with these systems.

One problem in the industry right now, of course, is that there are very few standards, something Chatrash is keenly aware of. “For the quantum industry to succeed, we need lots of startups like ourselves who are hyper-specialized in one particular area, because without companies who are hyper-specializing, we will not get economies of scale,” he said. “I think this whole full-stack story has to stop sooner or later. People need to start building an ecosystem of companies.”

Source : QuantrolOx uses machine learning to control qubits