Optimization of Quantum-dot Qubit Fabrication via Machine Learning

Abstract

Precise nanofabrication represents a critical challenge to developing semiconductor quantum-dot qubits for practical quantum computation. Here, we design and train a convolutional neural network to interpret in-line scanning electron micrographs and quantify qualitative features affecting device functionality. The high-throughput strategy is exemplified by optimizing a model lithographic process within a five-dimensional design space and by demonstrating a new approach to address lithographic proximity effects. The present results emphasize the benefits of machine learning for developing robust processes, shortening development cycles, and enforcing quality control during qubit fabrication.

References in corpus (5)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Improving neural networks by preventing co-adaptation of feature detectors
• Single-shot read-out of an individual electron spin in a quantum dot
• An addressable quantum dot qubit with fault-tolerant control fidelity
• A Reconfigurable Gate Architecture for Si/SiGe Quantum Dots

Cited by in corpus (7)

• Colloquium: Advances in automation of quantum dot devices control
• Toward Robust Autotuning of Noisy Quantum Dot Devices
• Neural networks for on-the-fly single-shot state classification
• Machine Learning-Assisted Manipulation and Readout of Molecular Spin Qubits
• Visual explanations of machine learning model estimating charge states in quantum dots
• Combining machine learning with physics: A framework for tracking and sorting multiple dark solitons
• QDFlow: A Python package for physics simulations of quantum dot devices