Yang, Seungmo Moon, Kyoung-Woong Ju, Tae-Seong Son, Jong Wan Kim, Taeyoon Jeong, Yeon Joo Park, Bae Ho Park, Sungkyun Hwang, Chanyong 2026-02-04T05:30:45Z 2026-02-04T05:30:45Z 2026-02-02 2026-01 1936-0851 https://pubs.kist.re.kr/handle/201004/154189 Recent advancements in spiking neural networks (SNNs) have drawn inspiration from the human brain’s distinctive capabilities, leading to significant impacts on various aspects of our lives and scientific endeavors. The development of hardware-based Spiking Neural Networks (H-SNNs) is a crucial aspect of the progress in neuromorphic computing, particularly in the context of on-chip implementations. H-SNNs involve the hardware realization of specific functionalities used in SNN algorithms, such as the leaky integrate-and-fire (LIF). One crucial functionality among these is neuronal homeostasis. However, although its importance in learning processes has been established, its hardware implementation at a single device remains unexplored until now. Here, we introduce a novel skyrmion-based spiking neuron device, which achieves hardware-level implementation of neuronal homeostasis. We have successfully demonstrated the modulation of a neuron’s threshold value through gate voltage application, a major step forward in demonstrating the feasibility of neuronal homeostasis in a singular hardware unit. We believe that our findings have significant implications, not only for the development of advanced Spiking Neural Network (SNN)-based on-chip devices but also for the potential application of skyrmions in electronic devices. English American Chemical Society Magnetic Skyrmion Neurons with Homeostasis for Spiking Neural Networks Article 10.1021/acsnano.5c07048 1 ACS Nano, v.20, no.1, pp.240 - 247 ACS Nano 20 1 240 247 Y scie 001651829500001 2-s2.0-105027310754 Chemistry, Multidisciplinary Chemistry, Physical Nanoscience & Nanotechnology Materials Science, Multidisciplinary Chemistry Science & Technology - Other Topics Materials Science Article skyrmion neuromorphic computing homeostasis spiking neural networks spintronics