A new experimental processor combines many of the attributes needed for practical quantum computing, but a full-scale implementation of the powerful technology remains far off on the horizon.

The demonstration quantum processor, described in a paper published online today in Science, was devised by a group of researchers at the National Institute of Standards and Technology (NIST) in Boulder, Colo. The device allows for the reliable application of multiple logic operations on a small set of quantum bits, or qubits, as well as the transport of information nearly a millimeter across the device by physically moving the qubits.

The data-bearing qubits in this processor are stored in ions, or charged atoms, trapped in place by electric fields. Information is encoded on the qubits by tweaking their quantum state, in this case the atoms' spin, with laser pulses.

The demonstration processor is notable for its stable combination of logic operations, somewhat akin to the AND and NOT gates of classical computers, with the movement of information around the processor, says lead study author Jonathan Home, a postdoctoral researcher at NIST.

In the past, Home says, movement of qubits has typically degraded the information encoded on them beyond the point of usefulness. "And so the major advance here is that we've now implemented methods that allow us to not only move the information around, but after we've done all that transport, we can process the information just as well as we would have done before doing the transport," he says.

To keep the atoms stable, Home and his colleagues employed a process known as sympathetic cooling. Magnesium ions trapped alongside beryllium ion qubits act as refrigerants—the key being that the two species of ions interact with very different wavelengths of light. So lasers can be used to cool the magnesium refrigerants, which in turn cool the beryllium ions, without affecting the quantum state of the latter. "What that means is we can shine that light and we don't kill our qubits that are stored in the beryllium," Home says.

The NIST team "has been working towards the goal of demonstrating that this architecture does indeed work, and now for the first time they combined all the ingredients into a single experiment," says Boris Blinov, a physicist who leads a trapped-ion quantum computing group at the University of Washington in Seattle.

"Of course, this is still a very small quantum information processor: only two qubit ions and two 'refrigerator' ions for the sympathetic cooling," Blinov says. "Scaling up to many more qubits will require some hard work."

Home readily acknowledges that the new processor is but a small-scale demonstration and that adding more qubits will be difficult. "It certainly looks like we have all the techniques needed to operate in a bigger device, but actually building and controlling a bigger device is a technological challenge," he says.

Even the new device, although reasonably reliable, lacks the fidelity desired of a practical quantum computer. The processor works roughly 95 percent of the time, Home says, but the goal is an operational accuracy rate of 99.99 percent. "So you can see there's a long way to go there," he says, "in order to get to where we want to be."