Investigating widespread correlated errors in superconducting qubit arrays

With the advent of large arrays of superconducting qubits, such as Google’s Sycamore processor, it is now possible to investigate the prevalence of correlated decoherence mechanisms that affect large numbers of qubits. Correlated errors are of particular importance with regard to implementing surface code error correction schemes, as errors are assumed to be independent or only weakly correlated. Control crosstalk and stray couplings are known to generate errors that are correlated, but the correlation is limited to a relatively small number of qubits. However, error mechanisms that affect large numbers of qubits simultaneously are problematic for error correction. In order to quantify the prevalence of such errors, we take rapid, simultaneous time-resolved measurements of error rates across large numbers of qubits on a single chip.