Race is on among tech firms to build a computing ‘oracle’

Google, IBM and Microsoft are competing to build a quantum computer, a device that will utilise quantum theory to take computing power to a whole new level.

Google, IBM and Microsoft are among those pedalling to the medal. Marcio Jose Sanchez / AP Photo

In ancient times, it would have been called an oracle – a source of instant insight on the most perplexing problems. Now, scientists are closing in on making a device capable of such feats.

Its name is as enigmatic as the source of its power: the quantum computer.

After decades of research, tech giants Google, IBM and Microsoft are among those racing to unveil the first quantum computer, which will solve in an instant problems that would tie up today’s computers for millennia.

From designing drugs to simulating reality and forecasting the future, quantum computers promise to turbo charge the scientific revolutions started by conventional computers.

The secret of their power lies in their exploitation of bizarre phenomena normally confined to the sub-atomic world but which scientists can now conjure to order.

Quantum theory is notorious for defying common sense.

Even one of its pioneers, the Nobel Prize-winning Danish theorist Niels Bohr, declared that anyone who isn’t shocked by its claims hasn’t understood it.

Quantum theory predicts, for example, that particles such as electrons have a property called spin, a rough analogy to a spinning ball.

But unlike the everyday notion where objects rotate clockwise or anti-clockwise, quantum theory implies that left to themselves, electrons rotate in both directions at once, in a “superposition of states”.

Only when we try to observe them do they take up one state.

Stranger still, particles can be made to interact in ways leaving them entangled with each other.

This connection means that if one particle is altered, all of its entangled partners undergo exactly the same change instantly regardless of how far apart they are.

These are not just fanciful ideas, either. The existence of these bizarre properties has been confirmed in lab experiments.

During the 1980s, theorists began to realise these strange phenomena could be used as the basis of the ultimate computer.

The idea is simple enough. Conventional computers work by processing on-off signals called bits, represented as 1s and 0s.

Quantum theory allows the creation of supercharged versions of bits, known as qubits. It does so through the phenomenon of superposition, which means qubits aren’t restricted to just 1s and 0s but can be mixtures of both.

Thus the equivalent of 1,000 bits processed by today’s computers becomes 1,000 qubits – every one of which is not just a 1 or a 0 but both. That means the number of states jumps from 1,000 to 2 raised to the power 1,000 – a number equal to 1 followed by more than 300 zeros.

For 30 years the challenge has been to find ways to exploit this huge boost in number-crunching power. That means overcoming the fact that qubits in superpositions are incredibly sensitive to outside disturbance.

Stray electromagnetic fields or heat can cause the superposition to collapse, turning qubits into ordinary bits. Called decoherence, it threatens the accuracy of quantum computation. In 2007, a small Canadian firm called D-Wave made headlines by claiming to have created the first practical quantum computer.

Experts questioned whether the company had solved the decoherence problem and pointed out that its device was useful only for specific types of problem.

Tests later revealed quantum effects alive and well inside D-Wave’s ultra-cold device. Its uses proved sufficient to convince Google and Lockheed Martin, and others, to pay US$10 million, or Dh36.7m for one.

The company is now rumoured to be about to unveil a 2,000-qubit device that is 1,000 times faster than its predecessors.

While impressive, this is still far short of what theory predicts, ­reviving questions about the extent to which D-Wave’s device is exploiting quantum effects to drive its performance, which sceptics claim can be achieved using standard computers in any case. Whatever the truth, D-Wave’s device has given impetus to the quest to build the first all-purpose quantum computer.

Last month, the journal Nature reported that Google has teams at three research centres working on a prototype they believe can overcome the problem of errors caused by decoherence.

Reports suggest the device is a hybrid of the D-Wave device and Google’s own technology, and may make its debut this year.

Meanwhile, Microsoft is pinning its hopes of winning the race on a discovery in particle physics.

A quantum computer may be faster than a conventional computer but its reliance on delicate qubits threatens to pepper its calculations with errors.

In 2012, a physicist at Delft University of Technology in the Netherlands announced the discovery of a new particle that could form a more robust qubit.

Known as a Majorana fermion, it can lurk inside molecule-thin wires of special material and act as a qubit, staying relatively safe from decoherence.

Microsoft has hired an international team of experts to turn this theoretical advantage into a practical reality. Building a quantum computer is one thing, programming it to solve real-world problems is another.

It takes specially designed algorithms to exploit the power of thousands of qubits working together, and there’s a shortage of them.

To speed their development, D-Wave has just unveiled a set of software design tools and a simulation of a quantum computer that allows anyone to write and run programs.

It is a move that is already paying off. D-Wave reports that biochemists have already used the software to speed up techniques used to trace evolutionary changes in human DNA.

Despite the recent advances sceptics insist the quest to build the ultimate computer still has a long way to go. Perhaps so, but backing for growing optimism comes from an unusual source: America’s code-breaking centre, the National Security Agency.

Documents leaked to The Washington Post by Edward Snowden show that the NSA is concerned about the effect of quantum computing on code-breaking.

Fearing other governments might beat it to the punch and begin decoding America’s top-secret communications, the NSA has started to build its own.

Its progress is not clear but given the power of the quantum computer you can be sure the first to build it outside the world of spooks will shout it from the rooftops.

Robert Matthews is visiting professor of science at Aston University, Birmingham. His book, Chancing It, is published in paperback this month.