Physicists from the University of California at Santa Barbara have developed as scheme to help quantum computers correct themselves.
The team from UC Santa Barbara will reportedly create self-correcting circuitries that leverage next-gen quantum computing technology. Quantum bits (or qubits, as they’re called) don’t like to retain their states. Qubits’ inabilities to “sit still,” in comparison to traditional circuitry which relies on either a yes/no or true/false state to store information, is one of the major hurdles scientists currently working on quantum computing must overcome.
“One of the biggest challenges in quantum computing is that qubits are inherently faulty. So if you store some information in them, they’ll forget it,” said Julian Kelly, a graduate student and co-lead author of the research paper. “It’s hard to process information if it disappears.”
To overcome this obstacle, the group of scientists have devised a system that essentially lets qubits take care of each other, acting as backups upon backups to ‘store’ the state that they’re supposed to be in. Using a grid system consisting of 9 qubits, the researchers aim to have the neighboring qubits safeguard each other’s information through repetitive error detection and correction.
Dubbed the “surface code,” this scheme of error detection and correction relies on parity information, or the measurement of change from the original data. In other words, the qubits compensate rather than replicate the states of their neighbors.
‘[You] can’t measure a quantum state, and expect it to still be quantum,” according to Rami Barends, a fellow researcher. By locking a qubit into a single state, you essentially take away its superpositioning power, he added.
To help maintain the “quantum-ness” of qubits, Kelly says all you need to do is extract just enough information to detect errors.
For now, the surface code system has only been proven to correct “bit-flip” errors, but the researchers are working on improving the system to correct “phase-flip” errors as well.
