# Atomic-level Characterisation of Quantum Computer Arrays by Machine Learning

## TL;DR

This paper presents a machine learning framework using CNNs trained on simulated STM images to precisely identify dopant atom counts and positions in silicon qubits, aiding large-scale quantum computer fabrication.

## Contribution

It introduces a high-accuracy, high-throughput method for atomic-level dopant characterization in silicon qubits using simulated data and neural networks.

## Key findings

- Achieved over 98% fidelity in dopant number and position identification.
- Trained on 100,000 simulated STM images with noise.
- Applicable for large-scale, high-precision quantum device fabrication.

## Abstract

Atomic level qubits in silicon are attractive candidates for large-scale quantum computing, however, their quantum properties and controllability are sensitive to details such as the number of donor atoms comprising a qubit and their precise location. This work combines machine learning techniques with million-atom simulations of scanning-tunnelling-microscope (STM) images of dopants to formulate a theoretical framework capable of determining the number of dopants at a particular qubit location and their positions with exact lattice-site precision. A convolutional neural network was trained on 100,000 simulated STM images, acquiring a characterisation fidelity (number and absolute donor positions) of above 98\% over a set of 17,600 test images including planar and blurring noise. The method established here will enable a high-precision post-fabrication characterisation of dopant qubits in silicon, with high-throughput potentially alleviating the requirements on the level of resource required for quantum-based characterisation, which may be otherwise a challenge in the context of large qubit arrays for universal quantum computing.