Quantum computing is on cusp of commercial breakthrough
A Cambridge start-up’s operating system could lead to the creation of a software market
Article by James Hurley | The Times
The world does not exist unless you look at it. The future influences the past. Everything that could possibly happen does, in an array of parallel realities. These are just three of the titular Six Impossible Things described by John Gribbin in his book on the baffling world of quantum physics. As bizarre as they sound, the “impossible things” are, in fact, serious attempts at explaining the mysteries of the subatomic world.
Steve Brierley, founder of Riverlane, is an expert on the subject, but he isn’t one for wondering about the mind-bending implications of quantum mechanics. Instead, he’s preoccupied with helping to make the science useful.
The 38-year-old chief executive of the quantum software start-up is “a pragmatist”. He’s also an optimist who refused to bow to colleagues who warned there was no future in quantum computing.
Riverlane, spun out of Cambridge University, revealed last week that it had completed a successful trial of an operating system that could pave the way for a software market to be established for quantum computers, opening up an array of commercial applications that may promise breakthroughs in areas including drug development.
A normal computer relies on “bits” that are either a 1 or a 0, or on or off, so is limited by binary options. Quantum computers use quantum bits, or qubits, that can be 0 or 1 as well as a “superposition” of 0 and 1 — in other words, on and off at the same time, the computing equivalent of Schrödinger’s cat.
An exponential increase in computing power is promised. Last year, Google claimed to have achieved so-called “quantum supremacy”, the term for when a quantum computer can outperform even the most powerful conventional computer for a specific task, a suggestion that opened up a feud with IBM, its quantum rival. Such a row would have been unthinkable even five years ago, according to Mr Brierley, who holds a PhD in quantum information from the University of York and is an expert in developing algorithms for quantum computers. In 2015, he was speaking at a quantum computing conference to the industry’s great and good. A poll asked delegates how long it might take to build a quantum computer of scale.
“They have all the top people in the field and what really shocked me at the time was that about a third of them voted that it was so hard to build the hardware that this would never happen.” Mr Brierley recalls it as a “crisis moment . . . Maybe I was developing algorithms for a computer that would never exist.”
He started researching the progress being made by companies, many of them based in Britain, developing the building blocks of quantum computers and was reassured that something resembling Moore’s law was in evidence; the principle stating that the number of transistors on a traditional microchip doubles roughly every two year.
“Research labs were getting twice as good at building small devices at least every one to two years. If you extrapolated that out, not only was one third of the audience wrong, a [working quantum computer] was going to happen much sooner than expected. The achievements of the last five years have shocked experts in the field. I don’t think anyone really felt that the progress would be so rapid.”
Yet commercial applications for quantum computing have remained elusive. One of the obstacles is that each computer has used a bespoke operating system. Riverlane’s product, Deltaflow.OS, is a step towards a universal system that would allow applications to be developed that could be used on multiple machines.
Deltaflow emerged from Riverlane’s own difficulties in running software on a local quantum computer. “It took six months to get a programme working. So we started building Deltaflow for ourselves because we needed a way to provide software for quantum computers.”
While the goal might be something akin to Windows for the quantum computer, a more appropriate analogy may be DOS, the operating system dominant in the PC market during the 1980s.
Mr Brierley likens the present position of the industry to the Homebrew Computer Club, an early hobbyist group established in 1975 in Menlo Park, California that helped to spark the rise of the PC. “You’ve got people wanting to understand the technology and try it out, but it being really quite hard to do so at the moment.
“When the first digital cameras were produced, they were rubbish compared to film cameras, but it’s a technology that completely changed how we take pictures and led to a whole new industry. Currently, quantum computers are not as good as a laptop. But new industries will be born out of this technology.”
Since it is hard to find algorithms that exploit the “superposition” advantage of a quantum system, the best bet for the emerging industry might be to focus on areas where quantum computers have the most natural advantage over their conventional counterparts. Since quantum mechanics can, for example, predict the behaviour of molecules as they approach a protein, a quantum computer could have a significant advantage in designing new materials, industrial processes or drugs, Mr Brierley believes.
“You can think of quantum computers as chemistry on steroids. Drugs bind to a protein, that is some molecule binding to some other molecule. That is a fundamentally quantum system. Quantum computers are essentially controllable quantum systems, which are very good at modelling other quantum systems.” That could mean thousands of different drugs being tested in the time it takes a lab to fabricate a single one.
Mr Brierley believes that we should expect to see some “meaningful impact” arising from the industry within five years.
It will solve problems that would otherwise never get solved. There are some computations that a conventional computer would never solve that a quantum computer can.
‘First came analogue, then came digital, quantum is what’s next’
The progress of Riverlane is the latest sign of Britain’s strength in quantum computing (James Hurley writes). About one in five of the world’s quantum hardware companies are based in the UK and all the various methods of building a quantum computer are in evidence.
In May the Cambridge-based Riverlane led a consortium that was awarded a £7.6 million grant to deploy a new operating system for quantum computing. Steve Brierley, its founder, argues that Britain’s promising position in the industry owes much to the decision by the government to back it, starting with investing in research more than a decade ago.
“A lot of these companies are spinouts from academic labs that have been very successful in the UK. That’s the reason Deltaflow has happened here — we can speak to loads of hardware labs. If you’re based in the US, you have to travel a long way to see the next type of quantum computer.”
Last week, it emerged that Rigetti, a California-based quantum start-up, would lead a £10 million consortium to create Britain’s first commercially available quantum computer. Mr Brierley believes that will “further strengthen the ecosystem here” and will provide a testbed for an emerging quantum industry.
M Squared Lasers, a Scottish laser developer and manufacturer, is providing “the shovels for the quantum goldrush”, according to Graeme Malcolm, its co-founder. Its lasers and control systems are used on quantum computers and Mr Malcolm suggests that Riverlane’s operating software is a “critical step in taking quantum computers out of the laboratory and into the commercial world, giving end-users the ability to test problems on different quantum hardware and systems”.
His maxim is that “first came analogue, then came digital, quantum is what’s next”. It will “revolutionise the economy and society alike”.
The government has invested hundreds of millions to seed the industry and, according to Mr Malcolm, now is not the time to let up: “We need to be advancing the UK’s quantum industry as a whole. From hardware to software, supply chains, skills and investment, in creating an entirely new sector we need all components to be moving in-step.”
Indeed, Mr Brierley believes that the government should go further and be ready to act as an early customer of commercial applications to help to demonstrate the potential of the technology. “This is going to take some time before it really emerges and if you want that to happen here, [state support] is going to be a big plus.”