x Abu Dhabi, UAESunday 21 January 2018

Quantum digits could dramatically speed up computers

Lockheed Martin is buying $10 million worth of computer hardware that could start the next technology revolution.

AAs technology deals go, it was hardly headline-grabbing: the giant US-based defence contractor Lockheed Martin is buying $10m (Dh36.7m) of computer hardware. But it may mark the start of the biggest revolution in computer technology since the invention of the microchip.

The hardware at the centre of the deal is said to be the world's first commercial quantum computer, a device that - in theory at least - could make today's computers seem as sluggish as an abacus.

Exploiting sub-atomic phenomena that defy classical physics, a quantum computer could solve myriad problems in one swoop, rather than cranking through each in turn. The result is an unimaginable increase in number-crunching power, with a single machine outgunning the combined power of every PC on Earth.

With those capabilities, $10m seems like a pretty amazing deal. The reality, of course, has a few caveats. Lockheed Martin has signed up to collaborate with D-Wave, the Canadian company that made the computer, to explore its powers over the coming years. And among scientists, that company is somewhat controversial. Based in the province of British Columbia, D-Wave made headlines four years ago by demonstrating what it said was a prototype quantum computer, which seemed a huge advance on anything else produced to that date.

The claim met with much scepticism, principally because it was not clear that D-Wave's machine really was exploiting the phenomena at the heart of a genuine quantum computer.

The central importance of these was first recognised by scientists in the 1980s. In conventional computers, data is processed in the form of 1s and 0s known as binary digits or "bits". So, for example, the number 9 in binary form becomes 1001, which thus requires four "bits" to represent it. In this form, data can be worked on by microchips, which churn through the bits one at a time.

Under the strange rules of the quantum world, however, it's possible to create bits that represent both 1 and 0 at the same time. The trick lies in a phenomenon called superposition, by which particles like electrons can exist in multiple states simultaneously.

A single electron can be put into a state so that it represents a mix of both 1 and 0 at the same time, creating the quantum equivalent of a binary digit, a "qubit". Qubits are thus an astonishingly efficient form of data: while four ordinary bits are needed to represent just the number 9, four qubits can represent all sixteen numbers (that is, 2 to the power 4) from 0 to 15 simultaneously.

Better still, quantum theory also allows whole groups of these qubits to be linked together - "entangled" - so their properties can be processed simultaneously. This leads to an awesome increase in number-crunching power. A quantum computer using just 100 qubits would be able to handle problems involving two to the power 100 different states in one go. That's around 10 to the power 30 - a one followed by 30 zeros, or more than the number of stars in the universe.

The applications of such power are endless, from designing life-saving drugs to predicting climate change and probing the origins of the cosmos. Frustratingly, though, scientists have to date had little success in realising such mind-boggling potential.

The biggest challenge lies in the delicate nature of qubits, which can lose their quantum properties all too easily, making them no better than ordinary bits.

So great is the effort required to ensure qubits keep their magic that, to date, the most complex problem solved using quantum computing is finding the prime factors of 15 (which are five and three, in case you don't have a quantum computer to hand).

D-Wave has tackled this delicate problem by looking for more robust types of qubit. They appear to have succeeded by using rings of so-called superconducting material that can be persuaded to carry current in two different directions at once. But while no one doubts that these materials exhibit quantum behaviour, it is less clear that the resulting qubits exploit the effects needed for a real quantum computer.

D-Wave has now answered some of its critics with research published recently in the journal Nature. This shows that the qubits really are acting in ways that can only be explained using quantum theory. Sceptics have been quick to point out, however, that this by itself is not enough: unless the qubits show both superposition and entanglement, they will never achieve the power of a full-blown quantum computer.

The company's scientists are convinced their qubits will turn out to have everything needed for a quantum computer. But in any case, if D-Wave's publicity is to be believed, the machine is still immensely powerful even if it is not the "real deal". The company says it is especially well-suited to solving certain types of real-life problem. These include "optimisation" - that is, finding really good solutions to problems with multiple answers, such as the quickest route between various cities.

Certainly Lockheed Martin is not the only major company impressed by its potential; Google has been experimenting with D-Wave's technology for several years. In 2009, the search giant announced that it had been using a D-Wave computer to sift through images and identify specific targets faster than possible with existing methods.

If D-Wave's machine lives up to its billing, its development will echo the story of another breakthrough technology: the aeroplane. When the Wright brothers announced the success of their first flight in 1903, they met with scepticism. For centuries people believed the only way to fly was to imitate the birds. The genius of the Wright brothers lay in showing that by taking a different tack it was possible to come up with a solution that, while lacking the elegance of nature, was still highly effective.

The dream of full quantum computing may never become a reality. But if the D-Wave machine takes us even a little way towards it, the consequences could be epochal.


Robert Matthews is visiting reader in science at Aston University, Birmingham, England