Crossbar Array Programming Using Piecewise-Constant Signals
https://doi.org/10.25682/NIISI.2025.3.0001
Abstract
To program a crossbar array, we need to adjust the resistor conductance using a limited number of control signals, which are voltages applied to the crossbar lines. Since the number of lines is significantly smaller than the number of resistors, this is a multi-step procedure. At each step, the conductances of the selected resistors are adjusted. The number of such resistors is no greater than the number of control signals. This inevitably changes the conductivity of some half-selected resistors, too. These unwanted changes must be compensated for. We examined a crossbar programming procedure using high-frequency piecewise-constant control signals. Our analysis involved a simple resistive element model. We demonstrated that an arbitrary (within known limits) conductance matrix can be programmed. At each step, a row or column of the crossbar array is generated or adjusted. We discussed the feasibility and convenience of such a procedure.
About the Authors
G. A. Beskhlebnova
НИЦ «Курчатовский институт» - НИИСИ, Москва
Russian Federation
Russian Federation
V. B. Kotov
НИЦ «Курчатовский институт» - НИИСИ, Москва
Russian Federation
Russian Federation
References
1. Kotov V.B., Beskhlebnova G.A. Specifics of Crossbar Resistor Arrays. // B. Kryzhanovsky et al. (Eds.). Advances in Neural Computation, Machine Learning, and Cognitive Research VI (NEUROINFORMATICS 2022). Studies in Computational Intelligence. Vol. 1064. Cham: Springer. 2023. PP. 292–304. https://doi.org/10.1007/978-3-031-19032-2_31.
2. Adamatzky A., Chua L. Memristor Networks. Springer International Publishing (2014).
3. Advances in Memristors, Memristive Devices and Systems. / Edited by S. Vaidyanathan and C. Volos. Springer International Publishing AG (2017).
4. Kim S. Ju, Kim S., Jang H.W. Competing memristors for brain-inspired computing. iScience 24, 101889, January 22, 2021.
5. Kotov V.B., Beskhlebnova G.A. Generation of the Conductivity Matrix. // B. Kryzhanovsky et al. (Eds.). Advances in Neural Computation, Machine Learning, and Cognitive Research V (NEUROINFORMATICS 2021). Studies in Computational Intelligence. Vol. 1008. Cham: Springer. 2022. PP. 276-284.
6. Surazhevsky I.A. at all. Noise-assisted persistence and recovery of memory state in amemristive spiking neuromorphic network.Chaos, Solitons and Fractals. 146 (2021). 110890.
7. Beskhlebnova G.A., Kotov V.B. The Variable Resistor Under a High-Frequency Signal. // B. Kryzhanovsky et al. (Eds.). Advances in Neural Computation, Machine Learning, and Cognitive Research VII (NEUROINFORMATICS 2023). Studies in Computational Intelligence. Vol. 1120. Springer Nature Switzerland AG. 2023. PP. 257–266. https://doi.org/10.1007/978-3-031-44865-2_28.
8. Kotov V.B., Beskhlebnova G.A. Use of High-Frequency Signals to Generate a Conductivity Matrix. // B. Kryzhanovsky et al. (Eds.). Advances in Neural Computation, Machine Learning, and Cognitive Research VIII (NEUROINFORMATICS 2024). Studies in Computational Intelligence. Vol. 1179. Springer Nature Switzerland AG. 2025. PP. 265-272.
9. Котов В.Б., Бесхлебнова Г.А. Поточечная запись информации в резисторную матрицу. // Труды НИИСИ РАН. Т. 14. №4. С. 33-40.
10. Kotov V.B., Yudkin F.A. Modeling and Characterization of Resistor Elements for Neuromorphic Systems. Optical Memory and Neural Networks (Information Optics). 2019, v.28, No.4, P. 271-282.
11. V.B. Kotov, Z. B. Sokhova. Two-frequency recording of information into a resistor array. // B. Kryzhanovsky et al. (Eds.). Advances in Neural Computation, Machine Learning, and Cognitive Research IX (NEUROINFORMATICS 2025). Studies in Computational Intelligence. Vol. 1241. Springer Nature Switzerland AG. 2026. PP. 463-476.
Review
For citations:
Beskhlebnova G.A., Kotov V.B. Crossbar Array Programming Using Piecewise-Constant Signals. SRISA Proceedings. 2025;15(3):9-16. (In Russ.) https://doi.org/10.25682/NIISI.2025.3.0001
Views:
53
JATS XML
Email this article 