Quantum computing

DAVID HAMBLING considers the possibilities, problems and implications of tech’s next big thing

Quantum computing is perhaps the weirdest technology ever devised, carrying out calculations via the almost-incomprehensible strangeness of quantum phenomena. The first quantum computers are now appearing, accompanied by a wave of hype. Is it real or fantasy? Alice-in-Wonderland quantum physics requires you to believe the impossible, and it’s not easy even for scientists to tell whether this next big thing is an illusion.

Quantum physics is founded on the observation that very small particles do not behave like familiar human-scale objects. They are fuzzy, and can seemingly be in several places at once, or teleport through barriers (‘quantum tunnelling’) or become connected so they communicate with each other at a distance (‘quantum entanglement’).

Physicist Niels Bohr said, “Anyone not shocked by quantum mechanics has not yet understood it,” and Einstein did not believe in it. But the reality of quantum physics is now beyond argument; the question is what to do with it.

Nobel laureate and bongo-playing eccentric Richard Feynman theorised in 1982 that quantum effects could drive a novel type of computer. Traditional digital computers are based on ‘bits’ set to either one or zero; a quantum computer is built around quantum bits or qubits which are one and zero at the same time. You cannot do much with two qubits, but things start to get interesting with a large number of them working together and quantum computing should excel at problems which involve working out vast numbers of combinations. If you take the many worlds view, quantum computing might be poetically seen as using millions of computers in parallel universes at once.

The idea is that quantum computers will achieve ‘quantum supremacy’ when they carry out feats that are impossible with other computers. Google’s quantumcomputing laboratory in Santa Barbara, California, claimed to have reached quantum supremacy in 2019 with a quantum device known as Sycamore with 53 qubits. In a paper in Nature, Google researchers claimed that Sycamore carried out a calculation in 200 seconds which would take a supercomputer over 10,000 years.

However, IBM rapidly counter-argued that one of their digital computers could solve the same mathematical problem, which involved simulating a quantum circuit, in two and a half days with a better algorithm. The argument has never been settled.

In 2020, researchers at the University of Science and Technology of China in Hefei claimed their 60-qubit quantum computer had solved a different problem, in a field known as boson sampling. This also took 200 seconds, rather than the two billion years it would take a traditional supercomputer.

By 2021, IBM announced a 127-qubit machine called Eagle, but have not yet made claims for its performance.

Quantum computers are causing excitement because they promise to carry out previously impossible calculations. Chemical engineers will be able to predict the properties of complex new molecules, identifying potent new drugs and other chemicals in hours rather than years. Quantum computing will tackle combinatorial problems in logistics and manufacturing, in particular in micro-electronics and artificial intelligence.

The money-makers are itching to get their hands on quantum computers for stock market prediction and modelling complex financial systems, both areas where even a modest advantage can quickly be leveraged into billions of dollars of profit.

The intelligence community is both excited and alarmed by quantum computing because of its potential for cryptography. Current digital encryption is secure because of the vast number of possible combinations that would need to be tried to unlock it. According to one estimate, standard RSA encryption would take 300 trillion years to crack with a conventional computer… or 10 seconds with a 4,000-qubit quantum computer.

The vast majority of Internet traffic, including financial transactions, is currently secured with RSA encryption and similar techniques. But a quantum code-cracker would change all that. Intelligence agencies are reportedly gathering huge archives of encrypted material now in the expectation of deciphering it in the near future. Everything which you hope is secret now may soon be laid bare.

Things may not, however, be quite so simple.

While some are already claiming quantum supremacy, nobody has yet used a quantum computer to tackle a meaningful real-world problem. Quantum computers are not yet identifying molecules, predicting markets or cracking codes. According to the best estimates, that will not happen until they reach the 1,000-qubit level, and even IBM have only just passed the 100 mark.

Digital electronic computers grew quickly from the 320-bits of the room-sized ENIAC in 1948 to the 1,600-bits of UNIVAC in 1951 to 16,000 bits of the IBM 650 in 1954, and on to the megabytes of PCs and gigabytes of current smartphones. But scaling up quantum computers is not just a matter of plugging in more qubits.

Quantum computers need to maintain a property called coherence, which is instantly destroyed by vibration, heat, radio waves or other interaction with the outside world. Maintaining coherence becomes progressively more difficult as the number of qubits gets scaled up. Some of the difficulty stems from the fact that the quantum effects require supercooling to within a fraction of a degree above absolute zero. Transferring electricity to quantum components heats them up, and packing more qubits together and increasing the number of connections makes the challenge progressively more difficult.

It is like building a house of cards: you easily get up to two or three levels, but getting up to 100 or more starts to look physically impossible.

The leaps required to bring quantum computing to anything like a useful scale may not be practically achievable. And while engineers are trying to figure them out, other types of computing may sneak up on the outside.

If you are baffled by quantum computing you are not alone. In the next few years it might change the world in ways we cannot imagine. Or it might remain a laboratory curiosity, the topic of endless debate. In quantum physics, waves are said to remain indeterminate until they ‘collapse’ and assume a definite state. Quantum computing itself still seems to be in that uncertain condition, somewhere between everything and nothing.