A.V. Bondarev1
1 Branch of Ufa University of Science and Technology (Kumertau, Russia)

1 Bondarevav@rambler.ru

The article is devoted to the study of the possibilities of using memristors in the development of new architectures of artificial intelligence systems based on modern nanoelectronic element base. A memristor is a promising class of electronic components with the ability to memorize resistance states, which opens up unique opportunities for solving complex problems typical of the fields of deep learning and cognitive modeling.

The author conducts a detailed comparative analysis of existing memristor models, evaluating their characteristics in terms of performance, power consumption, and scalability. Special attention is paid to the consideration of innovative approaches to the design of neuromorphic processors using membrane devices for efficient data storage and processing in analog circuits.

In addition, the problems of integrating memristors into existing integrated circuit manufacturing technologies have been investigated. Practical limitations and ways to overcome technical difficulties encountered in prototyping real devices have been considered.

The work also includes an experimental study of the functioning of memristor structures as part of the tested machine learning algorithms, demonstrating the potential benefits of switching to such devices. Recommendations are given for the further development of research in this area, which contributes to the expansion of scientific knowledge and stimulates the development of advanced solutions in the field of artificial intelligence.

The article is structured sequentially: from the introduction of the basic concepts and principles of the memristor to the description of specific examples of the implementation of neuromorphic networks and discussions of practical experimental results. Thus, this work is useful for specialists involved in the development of equipment for high-performance artificial intelligence, and serves as an important contribution to the formation of a modern understanding of the future directions of computer technology development.

Bondarev A.V. Analysis of memristor models for application in artificial intelligence systems based on nanoelectronic component base. Neurocomputers. 2026. V. 28. № 1. P. 46–57. DOI: https://doi.org/10.18127/j19998554-202601-05 (in Russian)

1. Bondarev A.V., Efanov V.N. Dynamic mode of the mathematical model of an electric multipole with memresistive branches in conditions of interval uncertainty. II International Scientific and Practical Conference «Actual problems of the energy complex: mining, production, transmission, processing and environmental protection». 16–17 April 2020. Moscow. P. 012013.
2. Bondarev V.A., Efanov V.N. Investigation of the robustness of nanoelectronic structures based on resonant tunneling elements. Proceedings of International Seminar on Electron Devices Design and Production. 2021. P. 9444533.
3. Memristor. Nanowerk [Elektronnyj resurs]. URL: https://www.nanowerk.com/memristor.php (data obrashcheniya: 01.05.23).
4. Babikov A.E., Bondarev A.V. Memristornye ustrojstva s ispol'zovaniem organicheskikh materialov. Sb. materialov XXI Mezhdunar. nauch.-praktich. konf. «Dostizheniya i perspektivy nauchnykh issledovanij molodezhi». Ufa: RITs UUNiT. 2023. S. 655–660. (in Russian)
5. Bondarev A.V., Efanov V.N. Analiz dinamicheskikh protsessov v nanoelektronnykh strukturakh na baze memrezistivnykh elementov. Izvestiya Samarskogo nauchnogo tsentra Rossijskoj akademii nauk. 2021. T. 23. № 2 (100). S. 91–97. (in Russian)
6. Bondarev A.V., Efanov V.N. Printsipy formirovaniya matematicheskoj modeli nanoelektronnykh komponentov kvantovykh vychislitel'nykh kompleksov s memrezistivnymi vetvyami. Sistemy upravleniya i informatsionnye tekhnologii. 2020. № 1 (79). S. 4–10. (in Russian)
7. Il'yasov A.I., Emel'yanov A.V., Nikiruj K.E. i dr. Massivy nanokompozitnykh krossbar-memristorov dlya realizatsii formal'nykh i impul'snykh nejromorfnykh system. Rossijskie nanotekhnologii. 2022. T. 17. № 1. S. 89–97. (in Russian)
8. Kiyatkin D.A., Bondarev A.V. Logicheskie elementy na osnove memristorov. Sb. materialov XXI Mezhdunar. nauch.-praktich. konf. «Dostizheniya i perspektivy nauchnykh issledovanij molodezhi». Ufa: RITs UUNiT. 2023. S. 691–694. (in Russian)
9. Gudkov A., Gogin A., Kik M. i dr. Memristory – novyj tip elementov rezistivnoj pamyati dlya nanoelektroniki. Elektronika NTB. 2014. № 9. S. 156–162. [Elektronnyj resurs]. URL: https://www.electronics.ru/journal/article/4756 (data obrashcheniya: 15.06.23). (in Russian)
10. Najdenov R.D., Bondarev A.V. Iskusstvennye nejrony na osnove nanorazmernogo memristornogo ustrojstva. Sb. materialov XXI Mezhdunar. nauch.-praktich. konf. «Dostizheniya i perspektivy nauchnykh issledovanij molodezhi». Ufa: RITs UUNiT. 2023. S. 694–699. (in Russian)
11. Ryabov I.R., Bondarev A.V. Operativnaya pamyat' mikroprotsessornykh ustrojstv na osnove memristorov. Sb. materialov XXI Mezhdunar. nauch.-praktich. konf. «Dostizheniya i perspektivy nauchnykh issledovanij molodezhi». Ufa: RITs UUNiT. 2023. S. 699–703. (in Russian)
12. Slepichko I.A., Bondarev A.V. Yachejki pamyati na osnove memristorov. Sb. materialov XXI Mezhdunar. nauch.-praktich. konf. «Dostizheniya i perspektivy nauchnykh issledovanij molodezhi». Ufa: RITs UUNiT. 2023. S. 710–714. (in Russian)
13. Dong J., Wang Y. Memristor-based artificial neural networks for machine learning. IEEE Transactions on Circuits and Systems I: Regular Papers. 2021. V. 68. № 10.
14. Chen L., Wu X., Zhang H. A survey of memristive neural network models in AI applications. ACM Computing Surveys. 2020. V. 53. № 3.
15. Rajendran B., Narayanan V., Karim F. Emerging devices enabling energy-efficient neuromorphic architectures. Frontiers in Neuroscience. 2021.