Quantum computing has long been the holy grail of physicists bent on transforming the way we communicate and how computers process information. It promises to be a vast improvement on existing computing technology for applications such as cryptography and will be able to calculate algorithms that computers can't handle at the moment.

At least, that's the promise. But in reality, it's a lot more challenging than it might sound.

Over recent months there has been several important advances towards quantum information processing – the basic building blocks for constructing a quantum computer – reported in major journals such as Nature and Science. And in June 2007, the quantum communication distance was broken when one half of an entangled photon pair was sent a distance of 144 km over the Canary Islands (see, Quantum communication breaks distance record, Cosmos Online).

But just how close are we to achieving quantum computing, and how important is it that we get there?

Hot on the trail

When people talk about how anything quantum works, it's not long before words like spooky, nonsensical, or wacky come into the conversation.

By definition, quantum is the singular of quanta, and is an indivisible unit of energy. Quantum mechanics looks at the weird properties inherent in nature that become apparent at atomic and sub-atomic scales, and quantum computing uses these processes to create something capable of forming logical operations.

Back in the 1999 the U.S. National Science Foundation (NSF) wrote that quantum information processing with the capacity of 10 qubits – the quantum equivalent of a classical computer 'bit' – would be possible within a decade, given the rate of progress of research in the quantum realm of physics.

We're not quite there yet, but research today is hot on the trail of quantum computing. Many small demonstrations of the theory are being achieved, as evidenced by the publication of a paper in the British journal Nature in July 2007 describing totally new ways to create the parts necessary for quantum computing.

"There is [currently] a huge amount of effort all over the world," says Andrew White, head of the quantum technology lab at the University of Queensland in Brisbane. But White acknowledges that quantum computers themselves are still at least 20 years away.

He's betting on a working quantum computer – as opposed to today's proof-of-principle demonstrations – by 2025. "I'm pretty sure it won't change the life of my students, but it might change the life of their students," he says.

Entangled pages

Quantum computing works in way analogous to classical computers, except that information is processed through qubits – quantum bits, rather than classical bits. Unlike classical bits, quantum bits can exist as either one state or another – such as 0 or 1 in binary language, the language used by computers – but also in a superposition of states. So a qubit may be a 1 or a 0, but its state is uncertain until it is measured.

Another advantage of quantum computers is that the information they use is contained not in the bits themselves, but in the correlation between the bits. The NSF uses the analogy of a book, where classical information is contained on the pages on the book, but quantum information in the relationship between the pages themselves. Thus if you remove a 'page' of the book, classical information is lost, but quantum information is unaffected.

What allows this to happen is a phenomenon occurring between quantum particles called entanglement. This property, described by Albert Einstein as 'spooky', is a non-local relationship between two particles such that an action undertaken on one particle – such as a measurement of its quantum state – affects the other particle, even if this particle isn't near its entangled partner. White describes it as "apparently nonsensical correlations" between particles.

It's entanglement which facilitates the ability of quantum information to 'teleport' — one of the more useful properties of quantum communication.

What's it all about?

All of this curious complexity allows quantum computers to process information exponentially faster than ordinary computers in certain applications, such as decrypting information.

"You can divide all problems into solvable and unsolvable. Of the solvable problems, some are efficiently computable and some are not," says White. Non-efficiently computable problems include factoring large numbers, the basis for cryptography behind, for example, online financial transactions.

In principle, quantum computers can solve these problems, hugely increasing the capacity for privacy and security, and also playing a role in detecting any attempt to 'break into' the code.

Quantum computing processes are also of interest to theoretical physicists like White, who are looking at fundamental problems in nature, such as how some semi-conductors work, or obtaining a full quantum simulation of molecules containing more than 30 atoms (caffeine, for example).

"If you have a full quantum model, you have a better understanding of a molecule works and reacts," says White. "[Qubits also represent] a controllable system that may shed more light on how quantum mechanics itself operates."

Making qubits

What's exciting scientists now are advances towards the components necessary for quantum computing: qubits and quantum gates. Over 20 different physical architectures are currently being investigated as contenders for components, the most successful being ions and photons.

Ions have been used to create the most qubits (eight), while photons have yielded the fastest quantum gates (quantum circuits analogous to classical logic gates in digital computers), and have been most comprehensively characterised, allowing researchers to detect crucial errors inherent in quantum processes.

Quantum information processing is prone to error because of the way components interact with the environment, known as decoherence. Up to seven qubits and several quantum gates may be needed to encode information reliably from one qubit, says White. To work out the factors for the number 15, for example (5 and 3), would take an enormous 4,000 qubits, each prone to error.

Another physical architecture being investigated for quantum information processing are neutral atoms. Scientists from the U.S. National Institute of Standards, in Gaithersburg, Maryland, led by Nobel laureate William Phillips, have developed a way to separate atoms in a supercool state and trap them in a lattice of lasers. Most of the atoms in the laser energy traps are linked with a partner in the opposite quantum state – with an electron spin up or down. This is the first time scientists have been able to pair up atoms in this way (see, New bits for qubits, Cosmos Online).

Flipping magnetic

The research, published in Nature in July 2007, is one step closer to making pairs of atoms capable of transferring quantum information. The next step is to scale up the process to work on arrays of atoms. Getting an array of atoms to act together in this way could create logical connections among data – the starting point for the logical processing necessary for computing to happen.

"Recent experiments have made progress towards this goal by demonstrating entanglement among an ensemble of atoms confined in an optical lattice," the researchers wrote. "Until now, however, there has been no demonstration of a key operation: controlled entanglement between atoms in isolated pairs... Our experiment realises proposals to use controlled exchange coupling in a system of neutral atoms."

Another study – conducted by researchers at institutions including John Hopkins University in Baltimore, Maryland, and published in U.S. journal Science in July – looked at how excitations in the magnetic 'flips' or fluctuations of the electron spins propagate across a ceramic substance.

While predominantly quantum mechanical, the research showed a previously unknown quantum 'order' that propagates along otherwise magnetically disordered material (low temperature ceramics). The propagation of the quantum order, known as phase coherence, can be controlled by introducing defects into the ceramic substrate or by changing its temperature – controlling the decoherence of the quantum order.

"We had two objectives," said study co-author Gabriel Aeppli, of the London Centre for Nanotechnology in Britain, "The first was to show that we could actually image the quantum order, which is sometimes referred to as 'phase coherence'. The second aim was to manipulate the distance over which it can be maintained."

This distance – and how sensitive it is to changes in temperature or chemical impurities in the material – can be essential in determining whether a material will have real-life applications, he said.

Reaching the chequered flag

The key factor to which physical architecture wins the day is the ability to scale upwards — creating enough qubits to complete a logic process — and the ability to withstand the effects of decoherence.

Whatever wins the day, White is excited about the final potential of quantum computer.

"It will change everything, in the same way current computers have changed everything. Think of MP3 players and mobile communication. In the early stages of classical computing, no-one predicted any of that would happen," says White.

Whatever the future holds for quantum computers, it may well be 20 more years before we see their final potential. It seems even technology that works like magic takes time to realise.

---

Add new comment