Another step on the long road to quantum computing

IBM has got error detection working on a segment of quantum computer.

The firm is working towards ‘surface code’ error correction for superconducting quantum computers, which need correction because entangled qubits (quantum bits) are prone to spontaneous errors. 

Surface code correction spreads quantum information across many qubits that only interact with immediate neighbours, suiting it to square arrays of qubits.

Qubits can exist in states including 1, 0 and superpositions of 1 and 0 (1 and 0 at the same time). In superposition, 1 and 0 have a phase relationship.

Possible errors include ‘bit-flip’: a qubit set to 1 spontaneously becoming 0 (or vice versa),  and ‘phase-flip’ where the relationship between 1 and 0 in superposition changes.

“Both types of errors must be detected in order for quantum error correction to function properly,” said IBM. “Others have addressed one error or the other, but not both at the same time.”

Direct measurement of a qubit destroys information, so the the lab has made part of a surface code with a square of qubits: two data qubits on opposing corners and two sensing ‘syndrome’ qubits opposing on the other corners.

“We employ ideas of entanglement and we can determine one of the characteristics of these data qubits through the measurement of one of these syndrome qubit, and get another aspect of the data qubits through the other syndrome qubit,” said IBM research manager Jerry Chow. “From that we can reconstruct what is the error [bit or phase-flip] that might have occurred to these data qubits.”

Scaled to a full surface code, this square would be part of a much larger lattice in which syndrome measurements would be made continually to determine what errors are occurring in the data qubits around them.

“We managed to show the ability to do a piece of the surface code,” said Chow.

How powerful will quantum computers be?

“If a quantum computer could be built with 50 qubits, there’s no combination of today’s TOP500 supercomputers that could successfully emulate it,” said IBM, although it might take two decades to build a useful quantum computer.

steve bush

from News http://ift.tt/1P6t2op
via Yuichun