Spintronic Artificial Neurons Showing Integrate-and-Fire Behavior with Reliable Cycling Operation
Source abstract
Abstract
The rich dynamics of magnetic materials makes them promising candidates for neural networks that, like the brain, take advantage of dynamical behaviors to efficiently compute. Here, we experimentally show that integrate-and-fire neurons can be achieved using a magnetic nanodevice consisting of a domain wall racetrack and magnetic tunnel junctions in a way that has reliable, continuous operation over many cycles. We demonstrate the domain propagation in the domain wall racetrack (integration), reading using a magnetic tunnel junction (fire), and reset as the domain is ejected from the racetrack with over 100 continuous cycles. Both the pulse amplitude and pulse number encoding are shown. By simulating a spiking neural network task, we benchmark the performance of the devices against an ideal leaky, integrate-and-fire neuron, showing that the spintronic neuron can match the performance of the ideal. These results achieve demonstration of reliable integrated-fire reset in domain wall-magnetic tunnel junction-based neuron devices for neuromorphic computing.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Advanced Memory and Neural Computing · Neural Networks and Reservoir Computing · Liquid Crystal Research Advancements
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Dept. of Electrical and Computer Engineering, University of Texas at Austin, Austin, Texas 78712, United States
Microelectronics Research Center, University of Texas at Austin, Austin, Texas 78758, United States
Funding
Basic Energy Sciences · Division of Graduate Education · Division of Electrical, Communications and Cyber Systems · University of Texas at Austin
References (44)
Show source-provided references
1. 10.1007/s00422-006-0068-610.1007/s00422-006-0068-6 ↗
2. 10.1109/TNN.2004.83272510.1109/TNN.2004.832725 ↗
3. 10.1109/LED.2018.285609210.1109/LED.2018.2856092 ↗
4. 10.1063/1.503192910.1063/1.5031929 ↗
5. 10.1002/adfm.20160474010.1002/adfm.201604740 ↗
6. 10.1016/j.nanoen.2020.10482810.1016/j.nanoen.2020.104828 ↗
7. 10.1002/adma.20180384910.1002/adma.201803849 ↗
8. 10.1002/aelm.20180086610.1002/aelm.201800866 ↗
9. 10.3389/fnins.2021.78669410.3389/fnins.2021.786694 ↗
10. 10.23919/VLSIC.2019.877810010.23919/VLSIC.2019.8778100 ↗
11. 10.1002/aelm.20220056310.1002/aelm.202200563 ↗
12. 10.1038/srep3151010.1038/srep31510 ↗
13. 10.3389/fnano.2022.102194310.3389/fnano.2022.1021943 ↗
14. 10.35848/1882-0786/ab7e0710.35848/1882-0786/ab7e07 ↗
15. 10.1021/acs.nanolett.9b0420010.1021/acs.nanolett.9b04200 ↗
16. 10.1109/TED.2021.311184610.1109/TED.2021.3111846 ↗
17. 10.1109/ACCESS.2022.319668810.1109/ACCESS.2022.3196688 ↗
18. 10.1021/acsnano.2c0974410.1021/acsnano.2c09744 ↗
19. 10.1126/science.adh341910.1126/science.adh3419 ↗
20. 10.1002/adma.20210367210.1002/adma.202103672 ↗
21. 10.1002/adfm.20210787010.1002/adfm.202107870 ↗
22. 10.1038/s42254-024-00729-w10.1038/s42254-024-00729-w ↗
23. 10.1038/s41928-020-0385-010.1038/s41928-020-0385-0 ↗
24. 10.1038/s41467-022-34309-210.1038/s41467-022-34309-2 ↗
25. 10.1038/s41467-023-39207-910.1038/s41467-023-39207-9 ↗
26. 10.1063/1.504245210.1063/1.5042452 ↗
27. 10.1109/LMAG.2021.306966610.1109/LMAG.2021.3069666 ↗
28. 10.1109/JXCDC.2022.322777410.1109/JXCDC.2022.3227774 ↗
29. 10.1063/5.003852110.1063/5.0038521 ↗
30. 10.1063/5.015221110.1063/5.0152211 ↗
31. Noh H. · Adv. Neural Inf. Process. Syst. · 2017 · 30 · 5101
32. 10.1109/TNNLS.2022.315252710.1109/TNNLS.2022.3152527 ↗
33. 10.1038/s41598-022-12555-010.1038/s41598-022-12555-0 ↗
34. 10.1088/1361-6463/aac60910.1088/1361-6463/aac609 ↗
35. 10.1038/s41928-023-01010-110.1038/s41928-023-01010-1 ↗
36. 10.1038/s41598-020-64878-510.1038/s41598-020-64878-5 ↗
37. 10.1016/S0925-2312(01)00658-010.1016/S0925-2312(01)00658-0 ↗
38. 10.1103/PhysRevLett.92.10720210.1103/PhysRevLett.92.107202 ↗
39. Paszke A. · Adv. Neural Information Process. Sys. · 2019 · 32 · 8024
40. 10.1016/j.jmmm.2007.12.02110.1016/j.jmmm.2007.12.021 ↗
41. Pehle, C.; Pedersen, J. E. Norse - A deep learning library for spiking neural networks. Zenodo, 2021. DOI: 10.5281/zenodo.4422024.
42. Xiao, H.; Rasul, K.; Vollgraf, R. Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms. arXiv (Machine Learning), Sept. 15, 2017, arXiv:1708.07747, ver. 2. DOI: 10.48550/arXiv.1708.07747.
43. 10.1162/neco.1996.8.3.64310.1162/neco.1996.8.3.643 ↗
44. Gulcehre, C.; Moczulski, M.; Denil, M.; Bengio, Y. Noisy Activation Functions. arXiv (Machine Learning), Apr. 3, 2016, arXiv:1603.00391, ver. 3. DOI: 10.48550/arXiv.1603.00391.