Quantum Leap

Teleportation, believe it or not, is key to the next generation of computers. But as Einstein showed, the world of the very small is very, very strange . . .

WHAT DOES teleportation have to do with computers? In this case, the answer has nothing to do with Star Trekand everything to do with one of the holy grails of computing: creating a quantum computer.

Last month, researchers reported in the journal Science that they got information from one atom to "teleport" to another - from one sealed container to another. It was immediately hailed as a spectacular demonstration that one of the key challenges on the road to quantum computing can be met: moving information between atoms so they can be harnessed to both store and transmit data, like infinitesimally small electronic parts in a computer.

According to many computer scientists and researchers, quantum computers - computers that work on the atomic level, utilising the very strange properties of physics in the realm of the very, very small - are where computing will inevitably end up in the future. When we get there, expect minute machines capable of incredible computing feats, with one solo quantum computer doing work that requires hundreds, even thousands, of machines working in tandem today.

Scientists argue that quantum computing is where we have to end up, because of the physical limitations that will impinge upon Moore's Law and transistors (see panel). As the intelligent building blocks of computers, transistors have served us well for decades, growing smaller and smarter about every 18 months. But they can't continue to get much smaller, or smarter, before they run into physical limitations on how they perform. So continued advancement in computing hardware means jumping down to the atomic and sub-atomic level.

The problem for researchers is that, as Albert Einstein proposed and as has since been proven, physics at the atomic level is like a strange, hallucinatory dreamworld.

Imagine a world in which a lamp could not just be switched on or switched off, but could be both on and off at the same time, and exist in every incremental state in between, simultaneously. Moreover, the moment you looked at the lamp and observed it, your observation would cause it to become fixed in one of the states, and then the lamp would be destroyed.

Very, very strange, yes; but that is how the quantum world operates. Although it sounds impossible and utterly bizarre, experiments with light - and the experiment with atomic teleportation - have proven this to be the case.

At the quantum level, particles are in every possible state simultaneously. Observing a particle influences what state it is seen to be in, and then destroys its quantum information as a knock-on effect. Why this happens is still debated.

How does this translate into a working computer? In a conventional, transistor-based computer, data is represented to the computer as either a one or a zero. Information is stored and processed as myriad strings of ones and zeros.

In a quantum computer, information can be stored in the form of ones, zeros and every possibility in between. Such a computer would be able to consider them all simultaneously and, when analysing data, calculate the results as all the possible variations that could happen. So a quantum computer can consider many different paths at the same time in order to reach a conclusion.

To achieve the same effect today - known as parallel computing - many, many machines and individual processors are needed to run calculations. A quantum computer theoretically should be able to do the same work in an instant, on a single machine.

The difficulty for researchers is working with the weird laws of quantum physics to create the quantum chips to run a quantum computer, figure out a mode of transport for the bits of quantum data (called "qubits", the smallest piece of quantum information) and design the memory needed to store data.

The biggest difficulty is uncertainty - the fact quantum particles can exist in different states simultaneously. But researchers say what seems to be quantum computing's biggest problem is actually its strength. Uncertainty, and all those multiple states, can be utilised to produce vast computing capabilities in a single computer. For any one problem, a quantum computer could view an astronomically large number of possibilities for computing an answer, but it would also be be able to select one computing path while simultaneously considering all those other possibilities. For this reason, quantum computers would be ultra-secure for encoding (encrypting) information: deciphering the one path a computer took to encode data, from an infinitesimal number of choices, would be extremely difficult.

But the flip side is also true - a quantum computer could crack more conventional codes very easily. Some recent experiments have indicated quantum computers may also be able to decrypt quantum encryption more easily than had been thought - though all of this remains in the realm of speculation.

For quantum computing, the importance of the successful experiment in teleportation is that researchers proved subatomic particles could be moved between two atoms in separate, sealed containers, and that this movement could be measured and verified (Einstein called this movement, in which doing something to one particle can change the status of another particle at a distance even though they are not in contact with each other, "spooky action at a distance"). That means, eventually, qubits of data could also be moved between atoms via a series of "quantum repeaters", forming the basis of information transport. And, they might also one day form the basis of an ultra-high-speed, quantum internet.

But so far, only tiny elements have been successfully built. A long research road lies ahead and quantum computers are probably still decades away.