Simulation of Memristive Crossbar Array Electrical Behavior in Neuromorphic Electronic Blocks

Мұқаба

Дәйексөз келтіру

Толық мәтін

Ашық рұқсат Ашық рұқсат
Рұқсат жабық Рұқсат берілді
Рұқсат жабық Тек жазылушылар үшін

Аннотация

A model and methodology for simulation of memristive crossbar arrays have been developed taking into account voltage drops on interconnections, the step of tuning the conductivity levels of memristive elements and the nonlinearity of their IV characteristics. The results of testing a spiking neural network in the inference mode in the problem of image recognition using the developed methodology of simulation taking into account the characteristics of experimentally manufactured memristor structures have been obtained.

Негізгі сөздер

Толық мәтін

Рұқсат жабық

Авторлар туралы

A. Dudkin

Saint Petersburg Electrotechnical University ETU “LETI”

Хат алмасуға жауапты Автор.
Email: rynenator@gmail.com
Ресей, St. Petersburg

E. Ryndin

Saint Petersburg Electrotechnical University ETU “LETI”

Email: rynenator@gmail.com
Ресей, St. Petersburg

N. Andreeva

Saint Petersburg Electrotechnical University ETU “LETI”

Email: rynenator@gmail.com
Ресей, St. Petersburg

Әдебиет тізімі

  1. Mutlu O., Ghose S., Gómez-Luna J., and Ausavarungnirun R. Processing data where it makes sense: Enabling in-memory computation // Proc. MICPRO, 2019. P. 28—41, https://doi.org/10.1016/j.micpro.2019.01.009.
  2. Li C., Hu M., Li Y. et al. Analogue signal and image processing with large memristor crossbars // Nature Electronics. 2018. V. 1. P. 52—59, https://doi.org/10.1038/s41928-017-0002-z.
  3. Li C., Belkin D., Li Y. et al. Efficient and self-adaptive in-situ learning in multilayer memristor neural networks // Nature Communication. 2018. V. 9. № 1. P. 2385—12385—8, https://doi.org/10.1038/s41467-018-04484-2.
  4. Li C., Wang Z., Rao M. et al. Long short-term memory networks in memristor crossbar arrays // Nature Machine Intelligence. 2019. V. 1. № 1. P. 49—57, https://doi.org/10.1038/s42256-018-0001-4.
  5. Hu M., Graves C.E., Li C. et al. Memristor-based analog computation and neural network classification with a dot product engine // Advanced Materials. 2018. V. 30. № 9. P. 1705914-1—1705914-5, https://doi.org/10.1002/adma.201705914.
  6. Sheridan P., Cai F., Du C. et al. Sparse coding with memristor networks // Nature Nanotechnology. 2017. V. 12. № 8. P. 784—789, https://doi.org/10.1038/nnano.2017.83.
  7. Du C., Cai F., Zidan M.A. et al. Reservoir computing using dynamic memristors for temporal information processing // Nature Communications. 2017. V. 8. 2204, https://doi.org/10.1038/s41467-017-02337-y.
  8. Marinella M.J., Agarwal S., Hsia A., Richter I., Jacobs-Gedrim R., Niroula J., Plimpton S.J., Ipek E., James C.D. Multiscale Co-Design Analysis of Energy, Latency, Area, and Accuracy of a ReRAM Analog Neural Training Accelerator // IEEE Journal on Emerging and Selected Topics in Circuits and Systems. 2017. P. 1—14, https://doi.org/10.1109/JETCAS.2018.2796379.
  9. Xu H., Sun Y., Zhu Y., Wang X., Qin G. Statistical Temperature Coefficient Distribution in Analog RRAM Array: Impact on Neuromorphic System and Mitigation Method. 2021, https://arxiv.org/pdf/2105.05534.
  10. Akhmetov Y., James A.P. Probabilistic Neural Network with Memristive Crossbar Circuits // IEEE International Symposium on Circuits and Systems (ISCAS). Sapporo, Japan. 2019. P. 1—5, https://doi.org/10.1109/ISCAS.2019.8702153.
  11. Gu P., Li B., Tang T. et al. Technological exploration of RRAM crossbar array for matrix-vector multiplication // The 20th Asia and South Pacific Design Automation Conference. Chiba, Japan. 2015. P. 106—111, https://doi.org/10.1109/ASPDAC.2015.7058989.
  12. Xu Q., Wang J., Yuan B. et al. Reliability-Driven Memristive Crossbar Design in Neuromorphic Computing Systems // IEEE Transactions on Automation Science and Engineering. 2023. V. 20. № 1. P. 74—87, https://doi.org/10.1109/TASE.2021.3125065.
  13. Thomas S.A., Vohra S.K., Kumar R., Sharma R., Das D.M. Analysis of Parasitics on CMOS based Memristor Crossbar Array for Neuromorphic Systems // IEEE International Midwest Symposium on Circuits and Systems (MWSCAS). Lansing, MI. USA. 2021. P. 309–312, https://doi.org/10.1109/MWSCAS47672.2021.9531867.
  14. Zhang L., Borggreve D., Vanselow F., Brederlow R. Impact of Parasitic Wire Resistance on Accuracy and Size of Resistive Crossbars // IEEE International Symposium on Circuits and Systems (ISCAS). Daegu, Korea. 2021. P. 1—5, https://doi.org/10.1109/ISCAS51556.2021.9401279.
  15. Kim H., Mahmoodi M.R., Nili H. 4K-memristor analog-grade passive crossbar circuit // Nature Communications. 2021. V. 12. 5198, https://doi.org/10.1038/s41467-021-25455-0
  16. Bayat F.M., Prezioso M., Chakrabarti B. et al. Implementation of multilayer perceptron network with highly uniform passive memristive crossbar circuits // Nature Communications. 2018. V. 9. 2331, https://doi.org/10.1038/s41467-018-04482-4
  17. Li Y., Ang K.-W. Hardware Implementation of Neuromorphic Computing Using Large‐Scale Memristor Crossbar Arrays // Advanced Intelligent Systems. 2020. V. 3. № 1. P. 2000137.
  18. Kim S., Hong K., Kim H., Kim M.‐H., Choi W. Overshoot‐Suppressed Memristor Array with AlN Oxygen Barrier for Low‐Power Operation in the Intelligent Neuromorphic Systems // Advanced Intelligent Systems. 2024. V. 2300797. P. 1—13, https://doi.org/10.1002/aisy.202300797.
  19. Deng L. The MNIST database of handwritten digit images for machine learning research // IEEE Signal Processing Magazine. 2012. V. 29. № 6. P. 141—142.

Қосымша файлдар

Қосымша файлдар
Әрекет
1. JATS XML
Жүктеу
2. Fig. 1. Memristive crossbar array: a - structure; b - implementation of matrix-vector multiplication.

Жүктеу (186KB)
Метадеректер
3. Fig. 2. Block diagram of the iterative algorithm for modeling of memristive crossbar arrays taking into account connection line resistances, nonlinearity of WACs and discreteness of the range of resistive states of memristive elements.

Жүктеу (194KB)
Метадеректер
4. Fig. 3. Schematic of the memristive crossbar array.

Жүктеу (224KB)
Метадеректер
5. Fig. 4. System of equations (5) in matrix form.

Жүктеу (162KB)
Метадеректер
6. Fig. 5. Experimental VAC of the memristor based on TiOx nanoscale film for 20 switching cycles. HRS and LRS are the high-resistance and low-resistance states of the memristor, respectively.

Жүктеу (84KB)
Метадеректер
7. Fig. 6. High impedance sections of the experimental VAC of the memristor based on TiOx film.

Жүктеу (72KB)
Метадеректер
8. Fig. 7. Family of VACs for different resistive states plotted by approximation of experimental data with a current step of 2σ.

Жүктеу (170KB)
Метадеректер
9. Fig. 8. Weight matrix of the software-pretrained impulse neural network.

Жүктеу (126KB)
Метадеректер
10. Fig. 9. Voltage drops on memristive crossbar array elements at different Rline resistances of the interconnect elements.

Жүктеу (133KB)
Метадеректер
11. Fig. 10. Voltage drops in the crossbar array at junction element resistance Rline = 10 ohms and different minimum values of RD memristive element resistances.

Жүктеу (146KB)
Метадеректер
12. Fig. 11. Example of iterative algorithm operation.

Жүктеу (92KB)
Метадеректер
13. Fig. 12. Comparison of currents flowing through the elements of a memristive crossbar array, with zero resistance of the interconnect lines (left) and 1 ohm resistance of the interconnect elements (right).

Жүктеу (177KB)
Метадеректер

© Russian Academy of Sciences, 2024