A machine learning approach for automated fine-tuning of semiconductor spin qubits

Abstract

While spin qubits based on gate-defined quantum dots have demonstrated very favorable properties for quantum computing, one remaining hurdle is the need to tune each of them into a good operating regime by adjusting the voltages applied to electrostatic gates. The automation of these tuning procedures is a necessary requirement for the operation of a quantum processor based on gate-defined quantum dots, which is yet to be fully addressed. We present an algorithm for the automated fine-tuning of quantum dots, and demonstrate its performance on a semiconductor singlet-triplet qubit in GaAs. The algorithm employs a Kalman filter based on Bayesian statistics to estimate the gradients of the target parameters as function of gate voltages, thus learning the system response. The algorithm's design is focused on the reduction of the number of required measurements. We experimentally demonstrate the ability to change the operation regime of the qubit within 3 to 5 iterations, corresponding to 10 to 15 minutes of lab-time.

Machine learning enables completely automatic tuning of a quantum device faster than human experts

“…Moreover, the transport features that indicate the device is tuned as a double quantum dot are time-consuming to measure and difficult to parametrise. Machine learning techniques and other automated approaches have been used for tuning quantum devices [5][6][7][8][9][10][11][12][13][14] . These techniques are limited to small regions of the device parameter space or require information about the device characteristics.…”

Deep reinforcement learning for efficient measurement of quantum devices

“…However, quantum dot devices are subject to variability, and many measurements are required to characterise each device and find the conditions for qubit operation. Machine learning has been used to automate the tuning of devices from scratch, known as super coarse tuning [34][35][36] , the identification of single or double quantum dot regimes, known as coarse tuning 37,38 , and the tuning of the inter-dot tunnel couplings and other device parameters, referred to as fine tuning [39][40][41] .…”

“…We have previously developed an efficient measurement algorithm for quantum dot devices combining a deep-generative model and an informationtheoretic approach 42 . Other approaches have developed classification tools that are used in conjunction with numerical optimisation routines to navigate quantum dot current maps 37,40,43 . These methods, however, fail when there are large areas in parameter space that do not exhibit transport features.…”

Computer-automated tuning procedures for semiconductor quantum dot arrays

“…One of the key challenges in scaling up spin qubits is developing the software tools necessary to keep pace with increasingly complex devices. To date, approaches to implementing automated control software during tune-up of semiconductor qubits include training neural networks to identify the state of a device 16 , experimentally realizing automated control procedures for tuning double quantum dot (DQD) devices into the single-electron regime 17 , and automatically tuning the interdot tunnel coupling in a DQD [18][19][20] .In this Letter, we use an image analysis toolbox developed at Sandia National Laboratories to accurately analyze charge stability diagrams acquired from a triple quantum dot (TQD) unit cell of a 9-dot linear array 13,21 . Computer automated analysis of charge stability diagrams performs the inversion of the device capacitance matrix and the establishment of 'virtual gates'.…”

“…One of the key challenges in scaling up spin qubits is developing the software tools necessary to keep pace with increasingly complex devices. To date, approaches to implementing automated control software during tune-up of semiconductor qubits include training neural networks to identify the state of a device 16 , experimentally realizing automated control procedures for tuning double quantum dot (DQD) devices into the single-electron regime 17 , and automatically tuning the interdot tunnel coupling in a DQD [18][19][20] .…”