Expedited Noise Spectroscopy of Transmon Qubits

Abstract There has been tremendous progress in the physical implementation of quantum protocols in recent times, bringing us closer than ever to realizing the promise of quantum computing. However, environmental noise continues to pose a crucial challenge to scaling up present‐day quantum processors. In the presence of uncontrollable noise sources, decoherence limits the qubits' ability to store information for long periods. Conventional noise spectroscopy protocols can characterize and model environmental noise but are usually resource‐intensive and lengthy. Moreover, the noise can vary in time, making the slow extraction futile as the profile cannot be harnessed to perform error mitigation or correction. Here, this challenge is addressed using a machine learning‐based methodology that outputs the noise associated with transmon qubits with minimal absolute error. The procedure involves implementing undemanding dynamical decoupling sequences to record coherence decays of the qubits and then predicting the underlying noise spectra with the help of a convolutional neural network pre‐trained on a synthetic dataset. While the protocol is virtually hardware‐agnostic, its effectiveness is validated using superconducting qubits available on the IBM Quantum platform. These rapidly obtained, yet accurate, noise spectra are further used to design bespoke dynamic decoupling sequences and perform time‐dependent noise spectroscopy.