Automated Atomic Silicon Quantum Dot Circuit Design via Deep Reinforcement Learning

Robert Lupoiu,Samuel S. H. Ng,Jonathan A. Fan,Konrad Walus

DOI: https://doi.org/10.48550/arXiv.2204.06288

2022-04-13

Abstract:Robust automated design tools are crucial for the proliferation of any computing technology. We introduce the first automated design tool for the silicon dangling bond quantum dot computing technology, which is an extremely versatile and flexible single-atom computing circuitry framework. The automated designer is capable of navigating the complex, hyperdimensional design spaces of arbitrarily sized design areas and truth tables by employing a tabula rasa double-deep Q-learning reinforcement learning algorithm. Robust policy convergence is demonstrated for a wide range of two-input, one-output logic circuits and a two-input, two-output half-adder, designed with an order of magnitude fewer SiDBs in several orders of magnitude less time than the only other half-adder demonstrated in the literature. We anticipate that reinforcement learning-based automated design tools will accelerate the development of the SiDB quantum dot computing technology, leading to its eventual adoption in specialized computing hardware.

Emerging Technologies,Mesoscale and Nanoscale Physics

What problem does this paper attempt to address?

This paper aims to solve the problem of the lack of automatic design tools in silicon dangling - bond quantum dot computing technology. Specifically, the author introduces an automated design tool based on Deep Reinforcement Learning (DRL) for the design of Silicon Dangling Bonds (SiDBs) quantum dot computing technology. This technology is a single - atom computing circuit framework with extremely high flexibility and versatility. However, due to the complexity and high - dimensional nature of its design space, manually designing these circuits is both time - consuming and difficult. Therefore, the method proposed in the paper can significantly accelerate the design process of SiDB logic gates and circuits, reduce the required time and resources, and at the same time improve the design efficiency and success rate. By using the Double - Deep Q - Learning algorithm, the paper solves the challenges encountered in the design process, such as the lack of gradient guidance for the objective function of the discrete logic truth table, the highly counter - intuitive nature of the optimization space, and the sparsity of the effective layout. This method can not only find successful strategies for design areas and truth table sizes of different scales, but also shows robustness to different random seeds, that is, it can converge to the same optimal design strategy regardless of the initial conditions. In addition, the paper also shows the practical application value of this automated design tool, for example, its role in creating the Bestagon gate library, which lays the foundation for higher - level SiDB design automation exploration.

Automated Atomic Silicon Quantum Dot Circuit Design via Deep Reinforcement Learning

Simulating Charged Defects in Silicon Dangling Bond Logic Systems to Evaluate Logic Robustness

Practical design and simulation of silicon-based quantum dot qubits

Differentiable Quantum Architecture Search for Quantum Reinforcement Learning

Design of high-performance entangling logic in silicon quantum dot systems with Bayesian optimization

Learning to Design Circuits

Machine Learning-Assisted Precision Manufacturing of Atom Qubits in Silicon

Challenges for Reinforcement Learning in Quantum Circuit Design

Chaotic diffusion of particles with finite mass in oscillating convection flows

A Reinforcement Learning Environment for Directed Quantum Circuit Synthesis

Hybrid discrete-continuous compilation of trapped-ion quantum circuits with deep reinforcement learning

Quantum Deep Dreaming: A Novel Approach for Quantum Circuit Design

Domain Knowledge-Based Automated Analog Circuit Design with Deep Reinforcement Learning

Quantum Compiling with Reinforcement Learning on a Superconducting Processor

Quarl: A Learning-Based Quantum Circuit Optimizer

The Need for Speed: Efficient Exact Simulation of Silicon Dangling Bond Logic

Variational Quantum Circuits for Deep Reinforcement Learning

QuickSim: Efficient and Accurate Physical Simulation of Silicon Dangling Bond Logic

Compiler for Distributed Quantum Computing: a Reinforcement Learning Approach

Attention-Based Deep Reinforcement Learning for Qubit Allocation in Modular Quantum Architectures