lunes, 24 de enero de 2011

In Pursuit of Qubits, Uniting Subatomic Particles by the Billions

In a step toward a generation of ultrafast computers, physicists have used bursts of radio waves to briefly create 10 billion quantum-entangled pairs of subatomic particles in silicon. The research offers a glimpse of a future computing world in which individual atomic nuclei store and retrieve data and single electrons shuttle it back and forth.
In a paper published this week in the journal Nature, a team led by the physicists John Morton of Oxford University and Kohei Itoh of Keio University describes an experiment in which they bombard a three-dimensional crystal with microwave and radio frequency pulses to create the entangled pairs.

This is one of a range of competing approaches to making qubits, the quantum computing equivalent of today’s transistors.

Transistors store information on the basis of whether they are on or off. In the experiment, qubits store information in the form of the orientation, or spin, of an atomic nucleus or an electron. The storage ability is dependent on entanglement, in which a change in one particle instantaneously affects another particle even if they are widely separated. The new approach has significant potential, scientists said, because it might permit quantum computer designers to exploit low-cost and easily manufacturable components and technologies now widely used in the consumer electronics industry.

“I think this is a very neat piece of work,” said Raymond Laflamme, a physicist at the University of Waterloo in Ontario, “but I think it’s important to see it as a piece of a big puzzle. Our mecca is to build a quantum computer that could have thousands of qubits; here we have only a few.”

Indeed, there is still disagreement over whether scientific or commercially useful quantum computers will ever be built. Until now, scientists have designed prototype quantum computers based on only a handful of qubits, too small a number to gain meaningful speed over conventional computers.

Unlike today’s binary computers, in which transistors can be in either an “on” or an “off” state, quantum computing exploits the notion of superposition, in which a qubit can be constructed to represent both a 1 and a zero state simultaneously.

The potential power of quantum computing comes from the possibility of performing a mathematical operation on both states simultaneously. In a two-qubit system it would be possible to compute on four values at once, in a three-qubit system on eight at once, in a four-qubit system on 16, and so on. As the number of qubits increases, potential processing power increases exponentially.

There is, of course, a catch. The mere act of measuring or observing a qubit can strip it of its computing potential. So researchers have used quantum entanglement — in which particles are linked so that measuring a property of one instantly reveals information about the other, no matter how far apart the two particles are — to extract information. But creating and maintaining qubits in entangled states has been tremendously challenging.

The new approach is based on a highly purified silicon isotope that is doped with phosphorus atoms. The research group was able to both create and measure vast numbers of quantum-entangled pairs of atomic nuclei and electrons when the crystal was cooled to about 3 degrees Kelvin. In the future the group hopes to produce the basis for a quantum computing system by moving the entangled electrons to simultaneously entangle them with a second nucleus.

“We would move the electron from the nuclear spin it is on to the neighboring nuclear spin,” said Dr. Morton. “That shifting step is what we really now need to show works while preserving entanglement.”

One of the principal advantages of the new silicon-based approach is that the group believes that it will be able to maintain the entangled state needed to preserve quantum information as long as several seconds, far longer than competing technologies which currently measure the persistence of entanglement for billionths of a second.

“To a member of the general public, that still sounds like a lousy time for a computer memory,” Dr. Morton said. “But for quantum information, the lifetime of a second is very exciting,” because there are ways to refresh the data.

The advance indicates that there is an impending convergence between the subatomic world of quantum computers and today’s classical microelectronic systems, which are reaching a level of miniaturization in which wires and devices are composed of just dozens or hundreds of atoms.

“This is on a single-nucleus scale, but it isn’t that far away from what is being used today,” said Stephanie Simmons, a graduate physics researcher at Oxford and the lead author of the paper. “There are two reasons people are taking a look at quantum computing. One is its power, but the other is that the size of silicon transistors are shrinking to the point where quantum effects are becoming important.”