ИСТИНА

Quantum memristors (QMs) have emerged as promising building blocks for neuromorphic quantum computing, offering the potential to combine the memory and nonlinear dynamics of classical memristors with quantum advantages such as high-dimensional state spaces and entanglement. Proposed realizations include various physical platforms such as superconducting circuits [1,2], quantum photonics [3-5], ion traps [6] and light-matter [7]. An experimental realization has been achieved in [5], where the authors have also demonstrated a numerical quantum advantage, comparing the performance of QMs versus classical memristors in an image recognition task. In this work, we propose and numerically validate a novel approach to define three coupled QMs on a single trapped 171Yb+ ion, enabling the operation of any two as a coupled pair at a given time. This architecture significantly reduces the number of ions required for complex neural architectures, paving the way for scalable multilayer quantum perceptrons. Through numerical simulations of the coupled system, we demonstrate that coupling preserves the memristive hysteresis behavior and enables signal transfer between QMs. Robustness analysis confirms that the hysteresis behavior remains stable under realistic experimental fluctuations in Rabi frequencies and pulse timing, with modern laser systems providing the necessary precision. We further evaluate the performance of QM-based models in a handwritten digit classification task using the MNIST dataset. Our results show that coupled QMs achieve recognition accuracies of 91–92%, matching the performance of ‘ideal’ memristors and uncoupled QMs. These findings validate the feasibility of coupled QMs for scalable, hardware-efficient quantum neural architectures and highlight their potential for advancing quantum-inspired machine learning. This study was supported by the Russian Science Foundation grant 24-12-00415, https://rscf.ru/project/24-12-00415/. References 1. Pfeiffer P., Egusquiza I. L., Di Ventra M., et al. Sci. Rep. 6 (1) 29507 (2016). 2. Salmilehto J., Deppe F., Di Ventra M., et al. Sci. Rep. 7 (1) 42044 (2017). 3. Sanz M., Lamata L., Solano E. APL Photonics 3 (8) 080801 (2018). 4. Gonzalez-Raya T., Lukens J. M., Céleri L. C., et al. Materials 13 (4) 864 (2020). 5. Spagnolo M., Morris J., Piacentini S., et al. Nat. Photonics 16 (4) 318–323 (2022). 6. Stremoukhov S. Yu., Forsh P. A., Khabarova K. Yu., Kolachevsky N. N. Entropy 25 (8) 1134 (2023). 7. Norambuena A., Torres F., Di Ventra M., et al. Phys. Rev. Appl. 17 (2) 024056 (2022).