Towards real-time quantum device calibration and drift mitigation

Quantum device performance depends on the quality of the underlying calibration. To get the best possible performance, control parameters must be carefully calibrated to maximize quantum gate fidelities and minimize the effects of noise. However, noise can change over time due to drifting conditions, undoing the benefits of good calibration. As such, strategies for keeping quantum devices stable and performant by tracking this drift are needed. In this talk, we develop low-latency drift mitigation protocols for this purpose, including policies based on real-time control parameter estimation and on reinforcement learning. We specifically develop strategies to compensate for drift in real time based on the outcomes of mid-circuit measurements, such as those utilized for quantum error detection. We present a variety of numerical analyses that examine the capabilities of these strategies to compensate for different drift processes.

SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525. SAND2024-14664A.