Research Fabrication of Integrated Neural Computing Circuits

集成神经计算电路的研究制造

• 批准号: 63850065
• 负责人: SAWADA Yasuji
• 金额: $ 10.69万
• 依托单位: TOHOKU UNIVERSITY
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Developmental Scientific Research
• 财政年份: 1988
• 资助国家: 日本
• 起止时间: 1988 至 1989
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-63850065/
• 关键词: Neuron Computer; Integration; AD-Converter; Neuron Device; Synapse Device; C MOS; Threshold; Accurate Function; CMOS; LSI化ニューロン回路; ヒドラ; 解離再集合体; ヒステリシスしきい値素子

An AD-converter was fabricated as the first step of research for integrated neuron networks. By constructing an AD-converter of integrated neural network one can study various fundamental problems underlying for the fabrication of general integrated neural networks except for the plasticity of the synapses. The designed circuit was a 4-bit system composed of 4 neurons and 12 synapses. A neuron includes 2 paris of CMOS transistors, and a synapse includes a pair of CMOS and a nMOS. The value of a synapse weight is determined by the geometry of the gate, L/W. The circuit has outside control lines to modify the characteristics in case the processes create some unexpected deviation. The pattern size was designed by 4 micron rule considering the various possible difficulties in the processes. The fabricated neuron showed a threshold characteristics with a sigmoidal curve as expected. The output of the synapse was proportional to the input with a small off-set. The off-set would be removed by changing the condition of the channel doping. The AD-converter with symmetric synapses proposed by Hopfield was found to operate with wrong results as the solution was trapped by a local minimum. A new AD-converter with desymmetrized synapses was found to operate always correctly. Fabrication of a new AD-converter is under way.

作为集成神经网络研究的第一步，研制了AD转换器。通过构造集成神经网络的AD转换器，人们可以研究除突触可塑性以外的一般集成神经网络的各种基本问题。所设计的电路是由4个神经元和12个突触组成的4位系统。神经元包括2个巴黎对CMOS晶体管，并且突触包括一对CMOS和nMOS。突触权重的值由门的几何形状L/W确定。该电路具有外部控制线，以在过程产生一些意外偏差的情况下修改特性。考虑到工艺中可能存在的各种困难，通过4微米规则设计图案尺寸。所制造的神经元显示出与预期的S形曲线的阈值特性。突触的输出与输入成正比，有一个小的偏移。通过改变沟道掺杂的条件可以消除这种偏移。由Hopfield提出的具有对称突触的AD转换器被发现以错误的结果操作，因为解决方案被局部最小值捕获。一个新的AD转换器与desymmetrized突触被发现总是正确地操作。一个新的AD转换器正在制造中。

项目成果

Yasuji.Sawada,Hirohumi.Yanai: "Integrator neurons for analogue neural networks" IEEE,Trans.on Circuit and Systems.

Yasuji.Sawada、Hirohumi.Yanai：“模拟神经网络的积分器神经元”IEEE，Trans.on Circuit and Systems。

Yasuji.Sawada,Hirohumi.Yanai: "Associated memory network composed of neurons with hysteretic property" Neural Network. (1990)

Yasuji.Sawada，Hirohumi.Yanai：“由具有迟滞特性的神经元组成的关联记忆网络”神经网络。

T.Itayama;O.Koizumi;Y.Sawada: Newral Networks. 1.Suppl.1. 475-475 (1988)

T.Itayama；O.Koizumi；Y.Sawada：Newral Networks。

H.Ando;Y.Sawada: J.Exp.Biol.

H.Ando；Y.Sawada：J.Exp.Biol。

H.Yanai;Y.Sawada: Newral Networks.

H.Yanai；Y.Sawada：Newral Networks。