A simpler way to test quantum computers

29 Sep 2008

Quantum computers could be a step closer thanks to a more classical approach to testing optical components.

Physicists in Canada have invented a new way of testing optical components that could someday be used to build quantum computers. They claim that their technique is much simpler than conventional tests because it uses standard laser light, rather than relying on the creation of photons in special quantum states (Sciencexpress 10.1126/science.1162086).

A quantum computer could, at least in principle, exploit the weird laws of quantum mechanics to vastly outperform classical computers on certain tasks. In such a computer, data would be input and stored in terms of quantum states — such as the polarization of individual photons. These data would be processed by devices that involve transitions in quantum systems, such as the absorption and emission of photons by a single atom.

But before any quantum computer can be made, these processing devices must be tested to ensure that they give the appropriate output state for a given input state — an exercise called quantum process tomography (QPT).

The conventional way of doing QPT is to systematically apply all possible inputs to a device and watch what comes out the other end. However, such input states are often very difficult to create in a reliable way — one of many reasons why physicists have yet to make a practical quantum computer.

Now, Alex Lvovsky and colleagues at the University of Calgary have come up with a way of doing QPT without the need to generate these tricky quantum states. It involves using much simpler, and more "classical", states of light from a conventional laser along with some high-powered mathematics.

The team demonstrated its technique using a continuous beam of coherent laser light that is shone through an electro-optical modulator (EOM) and then a polarizer. The EOM and polarizer can change both the amplitude and phase of the laser beam. At the microscopic level, this is a quantum process comprising a simultaneous absorption and phase shift of a quantum state.

To fully characterize this quantum process, the team adjusted the laser to create 11 different input states — which were beams with different amplitude and phase conditions. They then measured how these states were changed after passing through the EOM and polarizer.

This gave the group 11 different "views" of how the quantum process was affecting the laser light. Lvovsky and colleagues then used this information to construct a multi-dimensional set of equations that defines how any input quantum state is transformed into an output state. The technique is similar in some ways to medical imaging tomography, in which a 3D image is created from a series of 2D X-rays, for example.

Lvovsky likened the approach to that of a network analyser, an instrument that treats electronic circuits as "black boxes" by inputting a series of simple signals and measuring the outputs. "We study how the 'black box' processes simple coherent states, and this lets us know what it will do to any other state," he said.

Lvovsky told physicsworld.com that the team is now applying the technique to the study of quantum memory devices that store the information encoded in light. He also believes that the technique could be adapted for testing some non-optical quantum devices such as those based on the charge state of a tiny piece of superconductor.