X
Innovation

IBM shows off correctable quantum computing errors

IBM researchers will announce at the annual meeting of the American Physical Society in Boston today that they have established three new records for error correction in quantum computing.In a paper submitted on Feb 23rd for the conference, the researchers report a 95 per cent success rate with a two-qubit CNOT operation.
Written by Lucy Sherriff, Contributor

IBM researchers will announce at the annual meeting of the American Physical Society in Boston today that they have established three new records for error correction in quantum computing.

In a paper submitted on Feb 23rd for the conference, the researchers report a 95 per cent success rate with a two-qubit CNOT operation. This degree of accuracy is possible, they explain, because they have managed to maintain quantum coherence in their two-qubit computer for almost 10 microseconds. Per the IBM announcement: "These numbers are on the cusp of effective error correction schemes and greatly facilitate future multi-qubit experiments".

The other presentation (paper submitted on Feb 24th) will detail experiments with a three dimensional superconducting qubit, and idea IBM says it picked up from Yale University research. The new work from the computing firm shows that such an arrangement of qubits can be persuaded to maintain their quantum states for as long 100 microseconds. This length of time actually tips things over into the range of workable error correction, the company says.

Error correction is a huge part of quantum computing research, because so many things can disrupt carefully entangled qubit states: heat, magnetism, vibrations, materials defects. Any of these factors can introduce errors, which multiply quickly. But because an array of quantum states quickly decoheres – ie, loses its useful quantum mechanical properties and lapses into a simple classical state, it is hard to correct any errors that have been introduced. Extending the coherence time is key, then, to producing workable accurate quantum computers.

Editorial standards