{"title":"利用齐纳击穿进行神经形态计算的紧凑节能的LIF神经元设计","authors":"Faisal Bashir","doi":"10.1016/j.micrna.2025.208208","DOIUrl":null,"url":null,"abstract":"<div><div>In this paper, a single two terminal device (Zener diode) is used to construct a Leaky Integrate and Fire (LIF) neuron with significant improvement in energy efficiency, area and reduction in cost. Using calibrated 2D simulation, it has been confirmed that Zener diode based LIF neuron is able to imitate the neuron actions accurately. The Zener diode shows extremely sharper reverse breakdown voltage, the sharpness in the characteristics is responsible for achieving ultra-low energy per spike. The proposed Zener diode based LIF neuron needs only 86fJ/spike, which is 500 lower compared to recently reported Silicon on Insulator (SOI) based MOSFET, which needs 45pJ/spike of energy. Besides this, the proposed neuron design can be used to obtain neuron oscillation with different frequencies using an input current or capacitor values.</div></div>","PeriodicalId":100923,"journal":{"name":"Micro and Nanostructures","volume":"205 ","pages":"Article 208208"},"PeriodicalIF":3.0000,"publicationDate":"2025-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Compact and energy efficient design of LIF neuron leveraging Zener breakdown for neuromorphic computing\",\"authors\":\"Faisal Bashir\",\"doi\":\"10.1016/j.micrna.2025.208208\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>In this paper, a single two terminal device (Zener diode) is used to construct a Leaky Integrate and Fire (LIF) neuron with significant improvement in energy efficiency, area and reduction in cost. Using calibrated 2D simulation, it has been confirmed that Zener diode based LIF neuron is able to imitate the neuron actions accurately. The Zener diode shows extremely sharper reverse breakdown voltage, the sharpness in the characteristics is responsible for achieving ultra-low energy per spike. The proposed Zener diode based LIF neuron needs only 86fJ/spike, which is 500 lower compared to recently reported Silicon on Insulator (SOI) based MOSFET, which needs 45pJ/spike of energy. Besides this, the proposed neuron design can be used to obtain neuron oscillation with different frequencies using an input current or capacitor values.</div></div>\",\"PeriodicalId\":100923,\"journal\":{\"name\":\"Micro and Nanostructures\",\"volume\":\"205 \",\"pages\":\"Article 208208\"},\"PeriodicalIF\":3.0000,\"publicationDate\":\"2025-05-19\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Micro and Nanostructures\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2773012325001372\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"PHYSICS, CONDENSED MATTER\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Micro and Nanostructures","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2773012325001372","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"PHYSICS, CONDENSED MATTER","Score":null,"Total":0}
引用次数: 0
摘要
本文采用单端双端器件(齐纳二极管)构建LIF (Leaky Integrate and Fire)神经元,在能效、面积和成本上都有显著提高。通过校正后的二维仿真,证实了基于齐纳二极管的LIF神经元能够准确地模拟神经元的动作。齐纳二极管显示出极其尖锐的反向击穿电压,在特性的清晰度是负责实现超低的能量每尖峰。提出的基于Zener二极管的LIF神经元只需要86fJ/尖峰,这比最近报道的基于绝缘体上硅(SOI)的MOSFET低500,它需要45pJ/尖峰的能量。除此之外,所提出的神经元设计可以使用输入电流或电容值来获得不同频率的神经元振荡。
Compact and energy efficient design of LIF neuron leveraging Zener breakdown for neuromorphic computing
In this paper, a single two terminal device (Zener diode) is used to construct a Leaky Integrate and Fire (LIF) neuron with significant improvement in energy efficiency, area and reduction in cost. Using calibrated 2D simulation, it has been confirmed that Zener diode based LIF neuron is able to imitate the neuron actions accurately. The Zener diode shows extremely sharper reverse breakdown voltage, the sharpness in the characteristics is responsible for achieving ultra-low energy per spike. The proposed Zener diode based LIF neuron needs only 86fJ/spike, which is 500 lower compared to recently reported Silicon on Insulator (SOI) based MOSFET, which needs 45pJ/spike of energy. Besides this, the proposed neuron design can be used to obtain neuron oscillation with different frequencies using an input current or capacitor values.