PDA

View Full Version : It's the "Wiring" That's Tricky in Quantum Computing



popopot
05-08-2007, 01:15 PM
From: http://www.wired.com/science/discoveries/news/2007/05/quantumcoupling

While large-scale quantum computers remain in the domain of science fiction, a joint team from Japan announced Thursday that it has been able to take a small but crucial step in pursuit of this advanced goal.

NEC, the Institute of Physical and Chemical Research, or RIKEN, and the Japan Science and Technology Agency, published a paper in the May 4 issue of the journal Science, outlining the ability to "controllably couple qubits."

In classical computer science, bits -- or binary digits -- hold data encoded as ones and zeros. In quantum computing, data is measured in qubits, or quantum bits. As such, a qubit can have three possible states -- one, zero or a "superposition" of one and zero.

This unique property theoretically makes quantum computing able to solve large-scale calculations that would dwarf today's supercomputers. But qubits in isolation are not very useful. It's only when they can be connected to one another that large-scale processing becomes possible.

Controllable qubit coupling is analogous to the wiring of transistors on a circuit board. When qubits are coupled, they can affect one another -- thus acting as something like classical logic gates.

"This is an absolutely crucial set of tools for quantum computing to have -- the ability to controllably couple qubits," said Dmitri Averin, a physics professor at Stony Brook University, and an expert on quantum computing.

The NEC group is the second team in the world known to have controllably coupled qubits. John Clarke and his team at the University of California, Berkeley, published their results, albeit using a different technique, in December 2006, also in Science.

This latest study confirms the results of the Clarke group.

Until late last year, if you had qubit A and you needed it to be coupled to qubit B in order change the state of qubit B, you'd have to keep that link constantly active. This link -- the coupling -- is made possible by quantum entanglement. But keeping the link active is a problem because it will also change the state of qubit A -- when you only want to change the state of qubit B.

"Once this coherence is destroyed, you cannot do any quantum information processing," said Tsai Jawshen, an NEC fellow and a co-author of the paper. "That's the basic thing -- you have to let these guys remain in a quantum state."

For many years, scientists have been trying to figure out how to couple qubits for very short periods of time, just long enough to conduct a two-qubit operation, and to immediately shut it off once completed.

If controlling this coupling can be achieved, then larger computer logic operations should work.

However, in order to use this technique, and scale it up to a quantum computer with hundreds or thousands of qubits is another issue entirely, says Michele Mosca, the deputy director of the Institute for Quantum Computing at the University of Waterloo.

"Not only can you build them, but can you control it enough such that they won't fall apart too fast?" he said.

Indeed, no scientist is sure that such a large quantum computer wouldn't be subject to rapid quantum decoherence -- that is, the loss of the fundamental quantum mechanical properties that the computer would need in order to function.

Clarke said that when compared with his 2006 paper, the new study might be slightly better with respect to decoherence.

"The NEC version might have an advantage in that it can be used at the degeneracy point, (which is) a particular point where it's less sensitive to external noise sources, or something called flux noise," Clarke said. "At the degeneracy point, the device is less sensitive to decoherence."

In the NEC study, the team proved its ability to control qubit coupling in a three-step operation, Tsai said.

First, the team took a qubit A in superposition and a qubit B in either state zero or one. Next, they coupled the two qubits using a microwave focused on a third qubit, which entangled the other two. Nearly instantaneously, both qubits would be in superposition and the coupling would be turned off. Finally, the superposition for qubit A would remain -- preserving its initial quantum state.

So far, this is the only sequence that the team has been able to achieve. Dr. Tsai says the NEC group is working on a more complex, five-step procedure that will allow some basic logic to be carried out. He hopes it will be completed by the end of the year.

"We are very excited that everything works well as designed," he said. "It's wonderful. We're hoping that we can apply these things to something more useful."

lynx
05-08-2007, 11:58 PM
Oh, ffs, not this crap again.

Heisenberg's uncertainty principle tells us that we can't know enough about quantum particles to measure their complete state. If we can't measure the state, we can't know if we have influenced it. Consequently we have no way of knowing whether or not any action we might take has influenced any subsequent measurement.

The upshot is that any "result" we observe is random, so we can obtain no significant knowledge from it.

Schrodinger's cat has no bearing on the situation either, the cat may be alive or dead. The theory is that we only find out when we open the box, and therefore up to that point the cat is in an superimposed state.

Schrodinger forgot one thing. The cat knows.