Google has reached a quantum computer for error correction

Retracting error rates on quantum computers is sometimes better. More qubits, of course.

The quantum bits, or qubits, that make up a computer quantum are prone to errors that could render the calculation useless unless corrected. To reduce that error rate, scientists strive to build a computer that can correct its own errors. Such a machine would combine the potential of multiple fallible qubits into one, a better qubit, called a “logical qubit”, which can be used to make calculations.SN: 6/22/20).

Knowing now they have shown the keystone in the amount of error correction. Increasing the number of bits in the logic bit can make it less prone to error, researchers at Google report on February 22nd. nature.

In the future, the extent to which computers can solve even the most important problems of traditional computers (SN: 6/29/17). To build those huge quantum machines, researchers agree that they need to use error-correcting methods to reduce error rates. While scientists have previously demonstrated that they can detect and correct simple errors with minimal amounts of computing power, error correction is still in its infancy.SN: 10/4/21).

The new development does not mean that researchers are ready to build a fully error-corrected quantum computer, “however, it has been shown that it is possible for error to work fundamentally,” scientist Julian Kelly of Google Quantum AI said in a February 21 news briefing.

Logical qubits store redundant information in multiple physical qubits. That redundancy allows the computer to check if they have encountered any errors and fix them on the fly. Ideally, the greater the logic of the qubit, the greater the error should be. But if the original flaws are qubitta, adding more problems than it solves.

Using Google’s Sycamore quantum chip, the researchers investigated two different sizes of logical qubits, one of 17 qubits and the other of 49 qubits. After steady improvements were made to perfect the qubits that make up the original physical machine, the researchers made errors that still eluded them. Larger logic qubits had a lower error rate, about 2.9 percent error correction per round, compared with smaller logic qubits estimator of about 3.0 percent, the researchers found.

That small improvement suggests that scientists may eventually get stuck in a regime where error correction can begin to propagate errors. “It’s a bigger goal to achieve,” says scientist Andreas Wallraff of ETH Zurich, who was not involved with the research.

But the result is only on the cusp of showing that error correction is corrected as learned people climb. A computer simulation of a quantum computer implementation suggests that if the size of the qubit ratio were to increase more and more, its error would certainly get worse. In addition to correcting the original errors, qubits will be needed for scientists to really capitalize on the utility of error correction.

However, milestones in quantum computing are so difficult to achieve that they are treated like pole vaulting, Wallraff says. You only want to be opened to the courts hardly to him.