Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Spiking neural networks with fatigue spike-timing-dependent plasticity learning using hybrid memristor arrays

Nature Electronics (2026)Cite this article

Subjects

Abstract

Neuromorphic systems based on spike-timing-dependent plasticity offer energy-efficient learning but face limitations in terms of adapting to high-frequency inputs, restricting their effectiveness in processing complex temporal information. Synaptic fatigue dynamics, analogous to biological short-term plasticity, can increase the effectiveness, but this feature is difficult to efficiently incorporate in hardware. Here we report a hybrid architecture in which arrays of memristors with distinct dynamics are paired to create synaptic elements with short-term fatigue and long-term memory. The elements consist of an interfacial dynamic memristor with high uniformity and intrinsic fatigue behaviour coupled to a hafnia-based one-transistor–one-non-volatile memristor. The design enables a hardware-efficient implementation of fatigue spike-timing-dependent plasticity, enhancing the temporal learning capabilities of spiking neural networks. We show that the resulting neural network can be used for unsupervised online learning with high adaptability to both rate- and timing-coded spikes, high noise resilience and superior performance over conventional spike-timing-dependent plasticity approaches.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to the full article PDF.

USD 39.95

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Hybrid-memristor-array-based hardware SNN system.
Fig. 2: Short-term synaptic fatigue dynamics of the IDM.
Fig. 3: Hardware fatigue STDP synaptic cell based on the hybrid memristive element of IDM and 1T1R NVM devices.
Fig. 4: Experimental fatiguing STDP learning for correlation detection.
Fig. 5: Experimental fatiguing STDP learning for event image recognition.
Fig. 6: Audio recognition with hybrid-memristors-based SNN system under high-frequency noise.

Similar content being viewed by others

Data availability

Source data are provided with this paper. They are also available via Zenodo at https://doi.org/10.5281/zenodo.17501455 (ref. 61). Other data that support the findings of this study are available from the corresponding author upon reasonable request.

Code availability

The codes used for the event data generation algorithm are available via Zenodo at https://doi.org/10.5281/zenodo.17501455 (ref. 61). Other codes used for data plotting are available from the corresponding author upon reasonable request.

References

  1. Merolla, P. A. et al. A million spiking-neuron integrated circuit with a scalable communication network and interface. Science 345, 668–673 (2014).

    Article  Google Scholar 

  2. Qiao, N., Mostafa, H., Corradi, F., Osswald, M. & Stefanini, F. A reconfigurable on-line learning spiking neuromorphic processor comprising 256 neurons and 128K synapses. Front. Neurosci. 9, 141 (2015).

    Article  Google Scholar 

  3. Larimer, P. & Strowbridge, B. Timing is everything. Nature 448, 652–653 (2007).

    Article  Google Scholar 

  4. Roy, K., Jaiswal, A. & Panda, P. Towards spike-based machine intelligence with neuromorphic computing. Nature 575, 607–617 (2019).

    Article  Google Scholar 

  5. Chen, G. K. et al. A 4,096-neuron 1M-synapse 3.8-pJ/SOP spiking neural network with on-chip STDP learning and sparse weights in 10-nm FinFET CMOS. IEEE J. Solid-State Circuits 54, 992–1002 (2019).

    Article  Google Scholar 

  6. Prezioso, M. et al. Spike-timing-dependent plasticity learning of coincidence detection with passively integrated memristive circuits. Nat. Commun. 9, 5311 (2018).

    Article  Google Scholar 

  7. Pedretti, G. et al. Memristive neural network for on-line learning and tracking with brain-inspired spike timing dependent plasticity. Sci. Rep. 7, 5288 (2017).

    Article  Google Scholar 

  8. Boyn, S. et al. Learning through ferroelectric domain dynamics in solid-state synapses. Nat. Commun. 8, 14736 (2017).

    Article  Google Scholar 

  9. Stefano, A. et al. Unsupervised learning by spike timing dependent plasticity in phase change memory (PCM) synapses. Front. Neurosci. 10, 56 (2016).

    Google Scholar 

  10. Srinivasan, G., Sengupta, A. & Roy, K. Magnetic tunnel junction based long-term short-term stochastic synapse for a spiking neural network with on-chip STDP learning. Sci. Rep. 6, 29545 (2016).

    Article  Google Scholar 

  11. Kiselev, M. Rate coding vs. temporal coding—is optimum between? In Proc. 2016 International Joint Conference on Neural Networks (IJCNN) 1355–1359 (IEEE, 2016).

  12. Sboev, A. et al. Solving a classification task by spiking neural network with STDP based on rate and temporal input encoding. Math. Meth. Appl. Sci. 43, 7802–7814 (2020).

    Article  MathSciNet  Google Scholar 

  13. Moraitis, T. et al. Fatiguing STDP: learning from spike-timing codes in the presence of rate codes. Proc. IEEE 103, 1379–1397 (2015).

    Google Scholar 

  14. Caporale, N. & Dan, Y. Spike timing-dependent plasticity: a Hebbian learning rule. Annu. Rev. Neurosci. 31, 25–46 (2008).

    Article  Google Scholar 

  15. Rosenbaum, R., Rubin, J. & Doiron, B. Short term synaptic depression imposes a frequency dependent filter on synaptic information transfer. PLoS Comput. Biol. 8, e1002557 (2012).

    Article  MathSciNet  Google Scholar 

  16. Abrahamsson, T., Gustafsson, B. & Hanse, E. Synaptic fatigue at the naive perforant path-dentate granule cell synapse in the rat. J. Physiol. 569, 737–750 (2005).

    Article  Google Scholar 

  17. Zhang, H. et al. Synaptic fatigue is more pronounced in the APP/PS1 transgenic mouse model of Alzheimer’s disease. Curr. Alzheimer Res. 2, 137–140 (2005).

    Article  Google Scholar 

  18. Avian, H. & Korngreen, A. Short-term depression shapes information transmission in a constitutively active GABAergic synapse. Sci. Rep. 9, 18092 (2019).

    Article  Google Scholar 

  19. Wang, Z. et al. Resistive switching materials for information processing. Nat. Rev. Mater. 5, 173–195 (2020).

    Article  Google Scholar 

  20. Choi, S. et al. 3D-integrated multilayered physical reservoir array for learning and forecasting time-series information. Nat. Commun. 15, 2044 (2024).

    Article  Google Scholar 

  21. Jiang, R. et al. Habituation/fatigue behavior of a synapse memristor based on IGZO–HfO2 thin film. Sci. Rep. 7, 9354 (2017).

    Article  Google Scholar 

  22. Gutsche, A., Siegel, S., Zhang, J., Hambsch, S. & Dittmann, R. Exploring area-dependent Pr0.7Ca0.3MnO3-based memristive devices as synapses in spiking and artificial neural networks. Front. Neurosci. 15, 661261 (2021).

    Article  Google Scholar 

  23. Yang, Y. et al. Probing nanoscale oxygen ion motion in memristive systems. Nat. Commun. 8, 15173 (2017).

    Article  Google Scholar 

  24. Yang, Y. & Huang, R. Probing memristive switching in nanoionic devices. Nat. Electron. 1, 274–287 (2018).

    Article  Google Scholar 

  25. Gai, P. L. et al. In situ aberration corrected-transmission electron microscopy of magnesium oxide nanocatalysts for biodiesels. Catal. Lett. 132, 182–188 (2009).

    Article  Google Scholar 

  26. Cao, F. et al. Atomistic observation of structural evolution during magnesium oxide growth. J. Phys. Chem. C 120, 26873–26878 (2016).

    Article  Google Scholar 

  27. Lewis, T. J. & Wright, A. J. The electrical conductivity of magnesium oxide at low temperatures. J. Phys. D: Appl. Phys. 1, 441–447 (1968).

    Article  Google Scholar 

  28. Hays, D. C., Gila, B. P., Pearton, S. J. & Ren, F. Energy band offsets of dielectrics on InGaZnO4. Appl. Phys. Rev. 4, 733-530 (2017).

    Article  Google Scholar 

  29. Robertson, J. Electronic structure and band offsets of high-dielectric-constant gate oxides. MRS Bull. 27, 217–221 (2002).

    Article  Google Scholar 

  30. Yang, X. et al. Impact of insulator layer thickness on the performance of metal–MgO–ZnO tunneling diodes. Nano Res. 9, 1290–1299 (2016).

    Article  Google Scholar 

  31. Abhijeet, B., Kevin, D. & Rashmi, J. Deep-subthreshold Schottky barrier IGZO TFT for ultra low-power applications. Solid State Electron. Lett. 2, 59–66 (2020).

    Article  Google Scholar 

  32. Yeon, H. W. et al. Structural-relaxation-driven electron doping of amorphous oxide semiconductors by increasing the concentration of oxygen vacancies in shallow-donor states. NPG Asia Mater. 8, e250 (2016).

    Article  Google Scholar 

  33. Li, Y. et al. Hybrid-contact Schottky-barrier IGZO thin-film transistors with low barrier sensitivity and high stability. Appl. Phys. Lett. 126, 073502 (2025).

    Article  Google Scholar 

  34. Jeong, J. H. et al. Specific contact resistivity reduction in amorphous IGZO thin-film transistors through a TiN/IGTO heterogeneous interlayer. Sci. Rep. 14, 10953 (2024).

    Article  Google Scholar 

  35. He, J. et al. Defect self-compensation for high-mobility bilayer InGaZnO/In2O3 thin-film transistor. Adv. Electron. Mater. 5, 1900125 (2019).

    Article  Google Scholar 

  36. Wang, Q. & Holzwarth, N. Electronic structure of vacancy defects in MgO crystals. Phys. Rev. B 41, 3211 (1990).

    Article  Google Scholar 

  37. Runevall, O. & Sandberg, N. Self-diffusion in MgO—a density functional study. J. Phys. Condens. Matter 23, 345402 (2011).

    Article  Google Scholar 

  38. Cheng, C. et al. Oxygen-vacancy-ordering-induced metal-insulator transition in MgO single crystals. Results Phys. 36, 105452 (2022).

    Article  Google Scholar 

  39. Meux, A. et al. Oxygen vacancies effects in a-IGZO: formation mechanisms, hysteresis, and negative bias stress effects. Phys. Status Solidi A 214, 1600889 (2017).

    Article  Google Scholar 

  40. Song, H., Kang, G., Kang, Y. & Han, S. The nature of the oxygen vacancy in amorphous oxide semiconductors: shallow versus deep. Phys. Status Solidi B 256, 1800486 (2019).

    Article  Google Scholar 

  41. Hong, H. et al. Quantitative analysis of defect states in InGaZnO within 2 eV below the conduction band via photo-induced current transient spectroscopy. Sci. Rep. 13, 13407 (2023).

    Article  Google Scholar 

  42. Ye, F., Kiani, F., Huang, Y. & Xia, Q. Diffusive memristors with uniform and tunable relaxation time for spike generation in event-based pattern recognition. Adv. Mater. 35, 2204778 (2023).

    Article  Google Scholar 

  43. Roldán, B. et al. Variability in resistive memories. Adv. Intell. Syst. 5, 2200338 (2023).

    Article  Google Scholar 

  44. Chen, Z. et al. All-ferroelectric implementation of reservoir computing. Nat. Commun. 14, 3585 (2023).

    Article  Google Scholar 

  45. Chen, S. et al. Wafer-scale integration of two-dimensional materials in high-density memristive crossbar arrays for artificial neural networks. Nat. Electron. 3, 638–645 (2020).

    Article  Google Scholar 

  46. Kwon, K. C. et al. Memristive devices based on two-dimensional transition metal chalcogenides for neuromorphic computing. Nano-Micro Lett. 14, 58 (2022).

    Article  Google Scholar 

  47. Lee, K. et al. 3D stackable vertical-sensing electrochemical random-access memory using ion-permeable WS2 electrode for high-density neuromorphic systems. Adv. Funct. Mater. 34, 231380 (2024).

    Google Scholar 

  48. Stromatias, E., Soto, M., Serrano-Gotarredona, T. & Linares-Barranco, B. An event-driven classifier for spiking neural networks fed with synthetic or dynamic vision sensor data. Front. Neurosci. 11, 350 (2017).

    Article  Google Scholar 

  49. Massa, R., Marchisio, A., Martina, M. & Shafique, M. An efficient spiking neural network for recognizing gestures with a DVS camera on the Loihi neuromorphic processor. In Proc. 2020 International Joint Conference on Neural Networks (IJCNN) 1–9 (IEEE, 2020).

  50. Colonnier, F., Seeralan, A. & Zhu, L. Event-based visual sensing for human motion detection and classification at various distances. In Image and Video Technology 75–88 (Springer, 2023).

  51. Moon, J. et al. Temporal data classification and forecasting using a memristor-based reservoir computing system. Nat. Electron. 2, 480–487 (2019).

    Article  Google Scholar 

  52. Zhong, Y. et al. Dynamic memristor-based reservoir computing for high-efficiency temporal signal processing. Nat. Commun. 12, 408 (2021).

    Article  Google Scholar 

  53. Torrejon, J. et al. Neuromorphic computing with nanoscale spintronic oscillators. Nature 547, 428–431 (2017).

    Article  Google Scholar 

  54. Romera, M. et al. Vowel recognition with four coupled spin-torque nano-oscillators. Nature 563, 230–234 (2018).

    Article  Google Scholar 

  55. Le Prell, C. G. & Clavier, O. H. Effects of noise on speech recognition: challenges for communication by service members. Hear. Res. 349, 76–89 (2017).

    Article  Google Scholar 

  56. Zhou, Y., Liu, Y. & Niu, H. Perceptual characteristics of voice identification in noisy environments. Appl. Sci. 12, 12129 (2022).

    Article  Google Scholar 

  57. Grant, K. J. et al. Predicting neural deficits in sensorineural hearing loss from word recognition scores. Sci. Rep. 12, 8929 (2022).

    Article  Google Scholar 

  58. Ma, W. et al. High-frequency hearing loss is associated with anxiety and brain structural plasticity in older adults. Front. Aging Neurosci. 14, 821537 (2022).

    Article  Google Scholar 

  59. Cramer, B., Stradmann, Y., Schemmel, J. & Zenke, F. The Heidelberg spiking data sets for the systematic evaluation of spiking neural networks. IEEE Trans. Neural Netw. Learn. Syst. 33, 2744–2757 (2022).

  60. Wu, X. et al. Spatiotemporal audio feature extraction with dynamic memristor-based time-surface neurons. Sci. Adv. 10, eadl2767 (2024).

    Article  Google Scholar 

  61. Dang, B. Spiking neural networks with fatigue spike-timing-dependent plasticity learning using hybrid memristor arrays. Zenodo https://doi.org/10.5281/zenodo.17501455 (2025).

Download references

Acknowledgements

This work was supported by the National Key R&D Program of China (grant number 2023YFB4502200), Guangdong Provincial Key Laboratory of In-Memory Computing Chips (2024B1212020002), Shenzhen Science and Technology Program (JCYJ20241202125907011), National Natural Science Foundation of China (grant numbers 92164302 and 62406260), Beijing Natural Science Foundation (grant numbers L234026 and L257010) and the 111 Project (grant number B18001). This work has been supported by the New Cornerstone Science Foundation and Financial Support for Outstanding Scientific and Technological Innovation Talents Training Fund in Shenzhen.

Author information

Authors and Affiliations

  1. New Cornerstone Science Laboratory, Beijing Advanced Innovation Center for Integrated Circuits, School of Integrated Circuits, Peking University, Beijing, China

    Bingjie Dang, Teng Zhang, Keqin Liu, Ru Huang & Yuchao Yang

  2. Institute of Physics, Chinese Academy of Sciences, Beijing, China

    Fanqi Meng, Qinghua Zhang & Lin Gu

  3. School of Psychology & Cognitive Sciences, Peking University, Beijing, China

    Liutao Yu & Si Wu

  4. New Cornerstone Science Laboratory, Guangdong Provincial Key Laboratory of In-Memory Computing Chips, School of Electronic and Computer Engineering, Shenzhen Graduate School, Peking University, Shenzhen, China

    Yuchao Yang

  5. Center for Brain Inspired Intelligence, Chinese Institute for Brain Research (CIBR), Beijing, China

    Yuchao Yang

  6. Center for Brain Inspired Chips, Institute for Artificial Intelligence, Frontiers Science Center for Nano-optoelectronics, Peking University, Beijing, China

    Yuchao Yang

Authors
  1. Bingjie Dang
    View author publications

    Search author on:PubMed Google Scholar

  2. Teng Zhang
    View author publications

    Search author on:PubMed Google Scholar

  3. Fanqi Meng
    View author publications

    Search author on:PubMed Google Scholar

  4. Keqin Liu
    View author publications

    Search author on:PubMed Google Scholar

  5. Liutao Yu
    View author publications

    Search author on:PubMed Google Scholar

  6. Qinghua Zhang
    View author publications

    Search author on:PubMed Google Scholar

  7. Si Wu
    View author publications

    Search author on:PubMed Google Scholar

  8. Lin Gu
    View author publications

    Search author on:PubMed Google Scholar

  9. Ru Huang
    View author publications

    Search author on:PubMed Google Scholar

  10. Yuchao Yang
    View author publications

    Search author on:PubMed Google Scholar

Contributions

B.D. and Y.Y. designed the study and prepared the manuscript. B.D. fabricated the IDM devices and the fatigue STDP SNN test system, as well as performed the electrical measurements. T.Z. and B.D. prepared the 1T1R memristor array. B.D., Q.Z., F.M. and L.G. performed the in situ TEM and scanning TEM characterizations. B.D., L.Y. and S.W. demonstrated the fatigue STDP learning method. Y.Y. and R.H. directed all experimental research and supervised this work. All authors analysed the results and implications and commented on the manuscript at all stages.

Corresponding author

Correspondence to Yuchao Yang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Electronics thanks Xin Guo and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Supplementary Figs. 1–24, Tables 1–4 and Discussion.

Supplementary Video 1

In situ TEM observation of W/IGZO/MgO/W device dynamics under cyclic sweeps from 0 V to –3 V over 120 s. The movie is played at 5× speed.

Source data

Source Data Fig. 2

Data for Fig. 2b,c,g.

Source Data Fig. 3

Data for Fig. 3b,d–i.

Source Data Fig. 4

Data for Fig. 4a–e.

Source Data Fig. 5

Data for Fig. 5a–e.

Source Data Fig. 6

Data for Fig. 6b–f.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Dang, B., Zhang, T., Meng, F. et al. Spiking neural networks with fatigue spike-timing-dependent plasticity learning using hybrid memristor arrays. Nat Electron (2026). https://doi.org/10.1038/s41928-025-01554-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • DOI: https://doi.org/10.1038/s41928-025-01554-4

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing