It will be a long experimental haul before the great potential of quantum effects can routinely be exploited for technological ends. A sense of practical purpose among researchers will encourage progress.
When the citizens of Geneva cast their votes in the Swiss federal elections on 21 October, they could be confident that their ballots were safe — thanks to the rules of quantum mechanics. The poll results were sent down an optical fibre from the counting station to a government data centre, and their integrity was safeguarded by a quantum encryption key transmitted through the same fibre. Such a key promises to be 100% secure. It is composed of a stream of single photons that each take a random, unpredictable polarization state, and any attempts at tampering or eavesdropping will be noticed by the sender and receiver.
Quantum cryptography was high on the agenda at a meeting∗ last month on quantum information technology. Whether or not the exercise in Geneva was genuinely motivated by security concerns, it demonstrates, as a first public deployment of quantum cryptography, that the technique is ready to enter the commercial market for data encryption (G. Ribordy, id Quantique, Geneva; J. Dubois, Senetas, Melbourne). But a recurring theme of the meeting was the pressing requirement to identify other areas of practical application for quantum-information systems.
Another major prospect is quantum computing. Quantum computers will not simply be faster versions of the computers we have today. Rather, they will carry out tasks that are hard to tackle with any classical approach: for example, factoring large numbers and searching databases. Algorithms for such tasks have been available for more than a decade, and it is realizing the hardware that remains the main barrier to progress.
The building-block of a quantum computer is the qubit, a versatile version of the conventional bit. Like its classical counterpart, a qubit has two well-defined levels, '0' and '1'. But it also has the curious property that it can be in both states at the same time, occupying them with a certain probability. This phenomenon, known as 'superposition', in principle offers a powerful way to perform calculations because several logic operations can be carried out simultaneously. Another useful property of qubits is that they can be entangled — that is, their states can be prepared so that they are correlated to each other, even though each is unknown. For example, if one qubit happens to be in level 0 the other one will occupy level 1, and vice versa. The exact outcome becomes fixed as soon as either of the two is measured.
A wide range of qubit designs — based on atoms, ions, electrons, photons and even superconducting currents — was highlighted at the meeting. In most cases, at least two qubits can now be connected so that some sort of logic operation can be carried out. In one of the most advanced approaches, in which qubits take the form of ions trapped in an electromagnetic field, up to eight qubits have been entangled with each other. A new experimental development is the construction of a so-called Toffoli logic gate with ion qubits (R. Blatt, Univ. Innsbruck). Toffoli gates are a familiar concept in classical computation, but are the subject of renewed interest because they may offer a solution to error correction, a crucial consideration for quantum computers. However, they require three inputs and are therefore more difficult to realize than the two-qubit logic gates demonstrated so far.
The main problem with qubits is that their quantum states are fragile, and quickly leak away into the environment. This raises the scaling issue. Coupling just a few qubits together seems feasible. But as an increasing number of them are connected, more quantum leaks occur, so that information is quickly lost. Part of the solution may lie in using photons, relatively robust quantum entities, to channel quantum information between remote qubits, and experimental work is under way to construct such optical quantum connections. Instead of building a computer of say, 5,000 qubits, a more realistic goal may be to optically connect 1,000 quantum registers of just five qubits — one for storage, one for communication and three auxiliary qubits to ensure fault tolerance (A. Sorensen, Niels Bohr Inst., Copenhagen)1.
Small-scale quantum computers, designed to carry out a specific task, could be just a few years away. But a take-home message from the meeting was that more immediate applications of quantum technologies are urgently required to keep industrial partners interested (T. Spiller, Hewlett-Packard, Bristol). The 'killer application' for quantum information is not yet known, and more practical ideas need to be generated to kick-start a new market2. It is sobering to realize that the inventors of the transistor did not foresee the huge integrated-circuit industry that would develop; their first idea for a useful application of transistors was in hearing aids. What we need to bootstrap quantum-information technology, says Spiller, are quantum hearing aids.
When it comes to near-future technological applications, quantum communication, and quantum cryptography in particular, seems to be the best bet. The record distance over which a quantum key has been transmitted, both through an optical fibre and through free space, is about 150 km. But if quantum-communication technology is to be widely developed, it will be necessary to improve efficiency. At the meeting there was much talk of 'quantum repeaters' — pieces of hardware that can temporarily store and release photons without losing their quantum states, and that are seen as essential for the effective distribution of quantum information over large networks and distances. The experimental challenge to construct quantum repeaters is probably on a par with the challenge to generate practical qubits. So far, quantum memories, the basic element of a quantum repeater, have been made from ensembles of cold gaseous atoms. But a solid-state form will eventually be required: atomic ensembles of rare-earth ions, inserted in a nonlinear optical waveguide, are among the first candidates to be investigated (M. Staudt, Univ. Geneva)3.
The quantum future looks bright, although it will take a sustained experimental push before basic effects such as entanglement, inherent randomness and superposition can be exploited in real devices (practical or otherwise). But although quantum mechanics has been one of the most successful theories of the past century, nobody can confidently claim to understand why it works so well; for instance, how two entangled particles seem to communicate with each other at a distance, without any interaction, is beyond anybody's comprehension. There is a nagging feeling that we are missing something. A quantum-information industry may indeed be just around the corner, but its underlying principles remain largely mysterious.
∗QIPC 2007: International Conference on Quantum Information Processing and Communication, 15–19 October 2007, Barcelona, Spain.