Automated long-range compensation of an rf quantum dot sensor

TL;DR

This paper introduces an automated method for long-range compensation of charge sensors in quantum devices, enabling stable operation despite device-induced cross-talk, which is crucial for scalable quantum computing.

Contribution

It presents a Bayesian optimization-based algorithm that maintains sensor configuration over a wide voltage range, advancing fully automated quantum device operation.

Findings

Allows 1 V x 1 V voltage sweeps while maintaining sensor tuning

Reduces the need for manual sensor adjustments during device operation

Enhances scalability of quantum device architectures

Abstract

Charge sensing is a sensitive technique for probing quantum devices, of particular importance for spin qubit readout. To achieve good readout sensitivities, the proximity of the charge sensor to the device to be measured is a necessity. However, this proximity also means that the operation of the device affects, in turn, the sensor tuning and ultimately the readout sensitivity. We present an approach for compensating for this cross-talk effect allowing for the gate voltages of the measured device to be swept in a 1 V x 1 V window while maintaining a sensor configuration chosen by a Bayesian optimiser. Our algorithm is a key contribution to the suite of fully automated solutions required for the operation of large quantum device architectures.