Quantum mechanics is an extremely successful theory, but also a troubling one. For many years progress was made by concentrating on the obvious applications, and not worrying too much about the counterintuitive world view that quantum mechanics implies. More recently, however, the development of quantum-information theory has reversed this approach. If we take seriously what quantum mechanics seems to be telling us about the world, we can use this "quantum weirdness" to do apparently impossible things.

Probably the most famous application of quantum mechanics is the quantum computer, which is capable of performing calculations that are impossible with any classical device. At first the questions that quantum computers could tackle were rather esoteric, but in 1994 Peter Shor of AT&T Laboratories showed how a quantum computer could factor large numbers, thus rendering most modern cryptographic systems potentially obsolete.

In 1996 David Cory and co-workers at the Massachusetts Institute of Technology (MIT) showed how nuclear magnetic resonance (NMR) - a technique best known for its applications in medical imaging and in chemistry - could be used to build small quantum computers. NMR systems are easily controlled by the magnetic component of electromagnetic fields and are only weakly affected by decoherence, and so progress was extremely rapid. Within two years, several two-qubit computers had been developed, and simple algorithms had been implemented. The race was on to build bigger and better NMR quantum computers, and to use them for more interesting tasks.

The lead in this race has been held by several different research groups, but has now been decisively claimed by Isaac Chuang's group at Stanford University and IBM's Almaden Research Center in California. Chuang and co-workers have implemented the simplest example of Shor's quantum-factoring algorithm (L Vandersypen 2001 Nature 414 883).

In the April issue of Physics World, Jonathan Jones of Oxford University, UK, describes how Chuang's group factored the number 15 using only seven qubits.