Quantum memristors (QMs) have emerged as a promising frontier in neuromorphic computing and quantum information processing. In this paper, we propose a different methodology for defining three coupled ion-trap-based QMs on a single ${}^{171}\mathrm{Yb}^{+}$ ion, with the flexibility to operate any two of them as a coupled pair at a given time, opening possibilities for scalable multilayer quantum perceptrons which require fewer trapped ions for the same architectural complexity. Our study systematically evaluates the efficacy of coupled and noncoupled QM-based models in an image classification machine learning task. Numerical simulations reveal that coupling QMs preserves their performance quality, supporting the adoption of coupled QMs in scalable neural architectures. Our results demonstrate that QM-based models maintain stability in compact neural networks and achieve a high recognition accuracy, making them viable candidates for efficient neuromorphic computing systems.

Close

PrevNext

View figure in article

• 
• 
• 
• 
• 

• Atomic & molecular processes in external fields
• Electronic excitation & ionization
• Fine & hyperfine structure
• Laser applications
• Measurement-based quantum computing
• Memory
• Neuromorphic computing
• Open quantum systems & decoherence
• Quantum memories
• Quantum nondemolition measurement
• Artificial neural networks
• Hamiltonian systems
• Population dynamics
• Quantum complex networks
• Trapped ions
• Atom & ion trapping & guiding
• Laser techniques
• Machine learning
• Photoexcitation
• Photon counting
• Rabi model