CN112819148B - 基于浮栅晶体管的脉冲神经元网络 - Google Patents

基于浮栅晶体管的脉冲神经元网络 Download PDF

Info

Publication number
CN112819148B
CN112819148B CN202011638759.4A CN202011638759A CN112819148B CN 112819148 B CN112819148 B CN 112819148B CN 202011638759 A CN202011638759 A CN 202011638759A CN 112819148 B CN112819148 B CN 112819148B
Authority
CN
China
Prior art keywords
floating gate
pulse
gate transistor
input
electrode
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202011638759.4A
Other languages
English (en)
Other versions
CN112819148A (zh
Inventor
王青
陈静
吕迎欢
谢甜甜
赵瑞勇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Huali Microelectronics Corp
Shanghai Institute of Microsystem and Information Technology of CAS
Original Assignee
Shanghai Huali Microelectronics Corp
Shanghai Institute of Microsystem and Information Technology of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shanghai Huali Microelectronics Corp, Shanghai Institute of Microsystem and Information Technology of CAS filed Critical Shanghai Huali Microelectronics Corp
Priority to CN202011638759.4A priority Critical patent/CN112819148B/zh
Publication of CN112819148A publication Critical patent/CN112819148A/zh
Application granted granted Critical
Publication of CN112819148B publication Critical patent/CN112819148B/zh
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/06Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G06N3/063Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using electronic means
    • G06N3/065Analogue means
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/06Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G06N3/061Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
    • HELECTRICITY
    • H01ELECTRIC ELEMENTS
    • H01LSEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
    • H01L29/00Semiconductor devices specially adapted for rectifying, amplifying, oscillating or switching and having potential barriers; Capacitors or resistors having potential barriers, e.g. a PN-junction depletion layer or carrier concentration layer; Details of semiconductor bodies or of electrodes thereof ; Multistep manufacturing processes therefor
    • H01L29/66Types of semiconductor device ; Multistep manufacturing processes therefor
    • H01L29/68Types of semiconductor device ; Multistep manufacturing processes therefor controllable by only the electric current supplied, or only the electric potential applied, to an electrode which does not carry the current to be rectified, amplified or switched
    • H01L29/76Unipolar devices, e.g. field effect transistors
    • H01L29/772Field effect transistors
    • H01L29/78Field effect transistors with field effect produced by an insulated gate
    • H01L29/788Field effect transistors with field effect produced by an insulated gate with floating gate

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Theoretical Computer Science (AREA)
  • Neurology (AREA)
  • General Physics & Mathematics (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Molecular Biology (AREA)
  • Power Engineering (AREA)
  • General Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • Ceramic Engineering (AREA)
  • Thin Film Transistor (AREA)

Abstract

本发明提供了一种基于浮栅晶体管的脉冲神经元网络,包括多节点输入单元和脉冲产生单元:所述多节点输入单元包括一多输入端浮栅晶体管,多输入端浮栅晶体管的多个栅极输入端分别连接外部的多个仿生传感器输入信号,源极接地,漏极接脉冲产生单元的正极;脉冲产生单元包括一Mott忆阻器,Mott忆阻器的负极连接工作电压,正极连接晶体管的漏极,并作为所述脉冲神经元网络的脉冲输出端。本发明给出了一种全新的电子传入神经元实现架构。该架构面向硬件神经形态脉冲神经网络的应用,实现了模拟信号到脉冲信号的转换,具有结构简单、功能多、功耗低等优点,更加适应于脉冲神经网络。

Description

基于浮栅晶体管的脉冲神经元网络
技术领域
本发明涉及神经元网络领域,尤其涉及一种基于浮栅晶体管的脉冲神经元网络。
背景技术
脉冲神经网络作为下一代神经形态计算技术,是构建高能效存算一体数据处理中心的理想选择。为实现脉冲机制的感存算一体智能处理系统,需要构建高效的感知信息接口(生物学上称为传入神经)来建立脉冲数据处理中心与传感器之间的实时联系。然而现有技术采用CMOS构建的电子传入神经元,存在功耗高、电路结构和工艺复杂等问题,难以适用于新型神经形态脉冲神经网络。
发明内容
本发明所要解决的技术问题是,提供一种基于浮栅晶体管的脉冲神经元网络,解决功耗高、电路结构和工艺复杂等问题,适用于新型神经形态脉冲神经网络。
为了解决上述问题,本发明提供了一种基于浮栅晶体管的脉冲神经元网络,包括多节点输入单元和脉冲产生单元:所述多节点输入单元包括一多输入端浮栅晶体管,多输入端浮栅晶体管的多个栅极输入端分别连接外部的多个仿生传感器输入信号,源极接地,漏极接脉冲产生单元的正极;脉冲产生单元包括一Mott忆阻器,Mott忆阻器的负极连接工作电压,正极连接晶体管的漏极,并作为所述脉冲神经元网络的脉冲输出端。
本发明给出了一种全新的电子传入神经元实现架构。该架构面向硬件神经形态脉冲神经网络的应用,实现了模拟信号到脉冲信号的转换,具有结构简单、功能多、功耗低等优点,更加适应于脉冲神经网络。
附图说明
附图1所示是本发明一具体实施方式所述的基于浮栅晶体管的脉冲神经元网络的结构示意图。
附图2所示是本发明一具体实施方式所述的电子传入神经元实现架构图。
附图3A和3B所示是本发明一具体实施方式所述的浮栅晶体管的器件结构图。
具体实施方式
下面结合附图对本发明提供的基于浮栅晶体管的脉冲神经元网络的具体实施方式做详细说明。
附图1所示是本发明一具体实施方式所述的基于浮栅晶体管的脉冲神经元网络的结构示意图,包括多节点输入单元和脉冲产生单元:
所述多节点输入单元包括一多输入端浮栅晶体管,多输入端浮栅晶体管的多个栅极输入端分别连接外部的多个仿生传感器输入信号V01、V02、……V0N,源极接地,漏极接脉冲产生单元的正极。脉冲产生单元包括一Mott忆阻器Rm,Mott忆阻器的负极连接工作电压,正极连接晶体管的漏极,并作为所述脉冲神经元网络的脉冲输出端Ud(t)
多节点输入单元对外部仿生传感器输入信号V01~V0N进行信息整合,得到多输入端浮栅晶体管的浮栅电压VF,结合附图1所示的结构,该电压由下式联合给出
其中V0至VN是每一列浮栅晶体管的浮栅电压,是每个栅极节点输入电压的加权平均,C01~C0N是第一列浮栅晶体管的浮栅和栅极氧化物之间的寄生电容,对应的CN1~CNN是第N列浮栅晶体管的浮栅和栅极氧化物之间的寄生电容。对其进行加权平均即为多输入端浮栅晶体管的浮栅电压VF,C1~CN是各列浮栅晶体管的栅极氧化物和顶层硅之间的寄生电容。
当多输入端浮栅晶体管的浮栅电压VF小于阈值电压VT时,多输入端浮栅晶体管不导通;当浮栅电压VF达到多输入端浮栅晶体管的阈值电压VT时,多输入端浮栅晶体管开始导通。
多输入端浮栅晶体管导通后,晶体管在线性区间工作,与Mott忆阻器Rm形成电学串联。由电流流过的Mott忆阻器,其阻值会在高低阻态之间翻转,且翻转的频率与流过的电流成正比。因此脉冲神经元网络的脉冲输出端Ud(t)处的电平即随之上下波动,波动的频率与MIFG的翻转速率一致,形成振荡脉冲信号。
基于上述原理,本具体实施方式给出了一种全新的电子传入神经元实现架构,其架构图见附图2。该架构面向硬件神经形态脉冲神经网络的应用,实现了模拟信号到脉冲信号的转换,具有结构简单、功能多、功耗低等优点,更加适应于脉冲神经网络。
在一个具体实施方式中,所述多节点输入单元的由多输入端浮栅晶体管是22nm工艺节点的全耗尽SOI材料作为衬底的多输入端浮栅晶体管,其器件结构如附图3A和附图3B所示。所述晶体管包括衬底(Subtrate)、衬底表面的埋层氧化物(Buried Oxide)、以及埋层氧化表面的顶层硅,所述顶层硅通过掺杂形成源(Source)、漏(Drain)、以及源漏之间的采用薄膜硅(Thin Si-body)材料形成的导电沟道。导电沟道的表面设置栅极(Gate)以及栅极表面的浮栅(Floating Gate)。其优点在于可以利用全耗尽SOI材料的特点,在顶层硅上通过掺杂直接形成源、漏、以及导电沟道,并直接形成晶体管之间的串联,不需要额外制作导电隔离阱,因此是一种低成本高效率的选择方式。
以上所述仅是本发明的优选实施方式,应当指出,对于本技术领域的普通技术人员,在不脱离本发明原理的前提下,还可以做出若干改进和润饰,这些改进和润饰也应视为本发明的保护范围。

Claims (2)

1.一种基于浮栅晶体管的脉冲神经元网络,其特征在于,包括多节点输入单元和脉冲产生单元:
所述多节点输入单元包括一多输入端浮栅晶体管,多输入端浮栅晶体管的多个栅极输入端分别连接外部的多个仿生传感器输入信号,源极接地,漏极接脉冲产生单元的正极,所述多节点输入单元的多输入端浮栅晶体管是全耗尽SOI材料作为衬底的多输入端浮栅晶体管,所述晶体管包括衬底、衬底表面的埋层氧化物、以及埋层氧化表面的顶层硅,所述顶层硅通过掺杂形成源、漏、以及源漏之间的采用薄膜硅料形成的导电沟道;导电沟道的表面设置栅极以及栅极表面的浮栅,在顶层硅上通过掺杂直接形成源、漏、以及导电沟道,并直接形成晶体管之间的串联;
脉冲产生单元包括一Mott忆阻器,Mott忆阻器的负极连接工作电压,正极连接晶体管的漏极,并作为所述脉冲神经元网络的脉冲输出端。
2.根据权利要求1所述的基于浮栅晶体管的脉冲神经元网络,其特征在于,所述全耗尽SOI材料作为衬底的多输入端浮栅晶体管采用22nm节点工艺。
CN202011638759.4A 2020-12-31 2020-12-31 基于浮栅晶体管的脉冲神经元网络 Active CN112819148B (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011638759.4A CN112819148B (zh) 2020-12-31 2020-12-31 基于浮栅晶体管的脉冲神经元网络

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011638759.4A CN112819148B (zh) 2020-12-31 2020-12-31 基于浮栅晶体管的脉冲神经元网络

Publications (2)

Publication Number Publication Date
CN112819148A CN112819148A (zh) 2021-05-18
CN112819148B true CN112819148B (zh) 2024-08-06

Family

ID=75857615

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011638759.4A Active CN112819148B (zh) 2020-12-31 2020-12-31 基于浮栅晶体管的脉冲神经元网络

Country Status (1)

Country Link
CN (1) CN112819148B (zh)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110416086A (zh) * 2019-07-10 2019-11-05 复旦大学 一种fd-soi结构的半浮栅晶体管及其制备方法
CN111753976A (zh) * 2020-07-02 2020-10-09 西安交通大学 面向神经形态脉冲神经网络的电子传入神经元及实现方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103311294B (zh) * 2012-03-14 2016-09-21 中国科学院微电子研究所 鳍式场效应晶体管及其制造方法
CN105701541A (zh) * 2016-01-13 2016-06-22 哈尔滨工业大学深圳研究生院 一种基于忆阻器脉冲神经网络的电路结构
US10860923B2 (en) * 2016-12-20 2020-12-08 Samsung Electronics Co., Ltd. High-density neuromorphic computing element
CN106910773B (zh) * 2017-02-21 2019-08-20 南京大学 多栅极神经元晶体管及其制备方法和构成的神经网络

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110416086A (zh) * 2019-07-10 2019-11-05 复旦大学 一种fd-soi结构的半浮栅晶体管及其制备方法
CN111753976A (zh) * 2020-07-02 2020-10-09 西安交通大学 面向神经形态脉冲神经网络的电子传入神经元及实现方法

Also Published As

Publication number Publication date
CN112819148A (zh) 2021-05-18

Similar Documents

Publication Publication Date Title
US6956280B2 (en) Integrated circuit apparatus and neuro element
JP3438241B2 (ja) 半導体神経回路装置
US12026609B2 (en) Area and power efficient implementation of resistive processing units using complementary metal oxide semiconductor technology
JP3289748B2 (ja) 半導体装置
EP1866848A2 (en) Electronic synapse device
US20230186060A1 (en) Novel neuromorphic vision system
US20200125935A1 (en) Semiconductor device having neural network
US20220149200A1 (en) Neuron, neuromorphic system including the same
US12105986B2 (en) Devices, chips, and electronic equipment for computing-in-memory
CN114757345A (zh) 一种忆阻联想记忆神经网络电路及其控制方法
US11586899B2 (en) Neuromorphic device with oxygen scavenging gate
CN114792130A (zh) 一种具有泄漏-积分-发射功能的人工神经元
CN112819148B (zh) 基于浮栅晶体管的脉冲神经元网络
CN112396176B (zh) 一种硬件神经网络批归一化系统
US20190251430A1 (en) Mixed signal cmos rpu with digital weight storage
CN112836812B (zh) 基于浮栅晶体管的神经元网络
Diwan et al. Balanced ternary logic gates with memristors
CN115605026A (zh) 一种可直接实现权重差分的无源忆阻交叉阵列器件
CN112101533B (zh) 运算装置
JP2929909B2 (ja) 電界効果型トランジスタ
CN111384943B (zh) 一种柔性仿神经元电路及基于其的脉冲神经网络
임수환 Hardware-based Neural Networks using gated Schottky diodes
CN115994568A (zh) 基于二维材料浮栅晶体管的存算一体神经网络构建方法
US20220172034A1 (en) Neural network
KR20240154203A (ko) 심층 신경망 훈련용 전하 저장형 시냅스 장치 및 이의 구동 방법

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant