In early December, a team of Chinese scientists announced that they had designed a quantum computer that could perform a set of mathematical calculations (known as computations) nearly 100 trillion times faster than the world’s most advanced supercomputer. This announcement came just over a year after Google announced its own quantum breakthrough—it had built a system that could complete a complex computation in 200 seconds, compared with the nearly 10,000 years that a state-of-the-art classical supercomputer would have taken to perform the same task.
“The Google announcement was a tipping point,” says Roger McKinley, challenge director for quantum technologies at UK Research and Innovation, a government funding body. “It demonstrated feasibility and a justification to invest.”
The big question now is: how close are quantum computers to real-world uses?
To begin to understand the real-world potential of quantum computing requires, for most people, some explanation about how quantum computers work differently to traditional supercomputers. Conventional computers are built on ‘bits’, the smallest increments of data, which encode information in one of two states: zero or one. Quantum computers, by contrast, work with ‘qubits’, which, reflecting the mind-bending norms of quantum mechanical behaviour, can be in two states at once.
“In quantum, you are encoding information in quantum mechanical degrees of freedom, and the manipulation of information occurs in a continuously changing space,” says Chad Rigetti, founder of Rigetti, a startup that builds quantum integrated circuits used for quantum computers. “It’s like comparing north and south poles [classical computing] with a point anywhere on the earth [quantum computing], in terms of longitudinal and latitudinal variations.”
In a practical sense this means that quantum computers can crunch millions of calculations at the same time. The idea is not new—Paul Benioff, a US physicist, developed the concept decades ago—but theory is steadily becoming reality thanks to investments by tech giants, support from forward-thinking governments and the diverse efforts of deep-tech startups, mostly spin-outs from academic institutions.
Quantum computing technology is nascent but shows truly disruptive potential. By manipulating all possible states of quantum particles, quantum computers promise to compute much more efficiently and quickly than classical supercomputers. They can perform many tasks that are impossible with existing methods, holding particular promise in modelling complex processes—advancing our ability to discover new materials, forecast weather and climate, and design lifesaving drugs.
But there are still challenges to overcome as technologists in academia and the private sector seek to prove that their machines are stable and can work at scale.
Quantum spring
Alan Baratz, CEO of D-Wave Systems, a Canadian company that has released what it claims is the world’s first commercially available quantum computer, says that the technology has already reached a level of commercial viability. “We can now solve real business problems at scale.”
The company’s cloud-based quantum computing service, called Leap, costs USD2,000 per hour for access to query processing, and USD100 per hour for hybrid solver services, which combine classical algorithms and quantum resources. Volkswagen, one of D-Wave’s clients, cut waste in its painting process by 80 per cent by calculating a more efficient sequence for its manufacturing tools.
Chemicals and materials industries stand to be significant beneficiaries. “It might take a battery company 18 months to design and test a new electrode. With quantum, you could simulate material properties down to the quantum state and do it in two weeks,” says McKinley.
Quantum simulation could change how scientific research and development is organised, allowing more modelling and simulation work prior to “wet science” experiments.
Knowledge-based industries, like finance, could also leverage quantum power. Rigetti is applying quantum computing to power machine learning algorithms applied to financial datasets. “Finance is a highly computational and mathematical industry, and there are opportunities to insert a new form of computer hardware into workflows and routines in areas like large-scale portfolio optimisation,” says Rigetti.
UK bank Standard Chartered is among the participants in a GBP10m (USD13.3m) consortium set up by Rigetti to develop commercial quantum applications. The bank is exploring quantum computing for processes including simulating financial portfolios and exponentially accelerating the generation of market data. Spain’s CaixaBank, meanwhile, has used quantum-powered machine learning to classify credit risk.
Experts believe that machine learning, a popular branch of artificial intelligence, might only achieve its potential if turbo-charged with quantum computers. “Machine learning needs better processing power; there is an argument that you won’t get to real machine learning without quantum computing because we’ve underestimated how complex machine learning actually is,” says McKinley.
There are still big technical hurdles to overcome. Chief among them is creating enough qubits on which to conduct processing, because manipulating subatomic particles is a delicate process with any disturbance making them unstable. Error rates are high for some quantum computing systems, and the longer the processing period takes, the more likely that errors will arise.
There are many ways to build quantum systems—using superconducting qubits, trapped ions, diamond and silicon—and it remains unclear which is the most effective. This is slowing down investment.
“We have a VHS-Betamax problem,” says McKinley, referring to the widespread presence of two different hardware formats in videography prior to the DVD and streaming era. “There is more than one hardware approach and investors are not sure which to back. I take this as a healthy sign that this is a transformational domain, but some investors see it as a sign that it is still early, and they want to wait for the pack to slim down.”
Government funding will prove an important bridge. Forward-thinking agencies have seen the societal implications of quantum advantage. McKinley cites quantum processing as having a significant beneficial impact in imaging and sensing, which could have applications from medical imaging to autonomous and driverless systems and emissions monitoring.
He also points out the security implications of falling behind in quantum computing. The UK national programme was kicked off after a look at the “flash crash” of May 2010, a financial shock that briefly wiped hundreds of millions of dollars off the stock market. “Here, the speed of processing led to instability in financial trading. In a global economy, once one company or country has new capabilities there will be implications”.
The UK government has invested around GBP1bn (USD1.3bn) in quantum computing, including GBP120m in four academic-industry hubs developing quantum approaches to sensors and positioning, time and imaging. McKinley thinks that the UK is a frontrunner thanks to its strengths in physics and engineering and, paradoxically, the space for startups and academic spinouts created by the lack of UK-based tech giants. “Because we haven’t got the mighty oaks, there is a lot of rain and sun getting to the forest floor. We’re not in the shadow of a Google or Microsoft.”
But the race is on as other governments invest their own resources. The US government recently announced USD1bn of new funding for AI and quantum computing research hubs to explore a range of applications from ocean science to high-energy physics. The EU has a EUR1bn (USD1.22bn) ten-year Quantum Technologies Flagship initiative. Canada has invested around USD1bn over the decade to 2019 and the Chinese government is reportedly building a USD10bn National Laboratory for Quantum Information Sciences.
One thing is clear: no-one wants to be left behind in the race for quantum advantage.
