Google uses error-correction data to keep quantum processor calibrated

Google has developed a method to recalibrate a quantum processor using data from error correction, solving a drift problem that affects superconducting qubits during long calculations.

This article was drafted with AI assistance from multiple sources and was reviewed and approved by a human editor before publication.

Google has found a way to keep a quantum processor calibrated during calculations by using data from the error-correction process, according to a company announcement on July 10. The technique addresses a challenge specific to hardware such as superconducting qubits, which require regular calibration to maintain low error rates.

Calibration is a process that tests different frequencies and amplitudes of microwave pulses to find the settings that produce the lowest error rates. For superconducting qubits, subtle variations among individual qubits make calibration necessary. However, standard calibration cannot be performed while a calculation is running, so drift in the qubits' properties can degrade performance over time, especially for long and complicated algorithms.

Google's approach uses the same data that is collected for error correction to continuously recalibrate the processor. This eliminates the need to pause calculations for separate calibration routines. The company says the method works with error-corrected logical qubits and can adjust for drift without interrupting the computation.

The breakthrough is one of many less prominent challenges that must be solved before practical quantum computing becomes viable. Other major issues include producing enough high-quality hardware qubits to form error-corrected logical qubits and generating the states needed for universal computation.

Google did not specify when the technique would be integrated into its quantum processors or provide performance benchmarks. The research was published on July 10.