{"title":"Reservoir Computing Utilizing a Complementary Combination of n- and p-Channel FeFETs","authors":"Rikuo Suzuki;Kasidit Toprasertpong;Ryosho Nakane;Eishin Nako;Mitsuru Takenaka;Shinichi Takagi","doi":"10.1109/LED.2024.3435422","DOIUrl":null,"url":null,"abstract":"We investigate the potential of physical reservoir computing utilizing a combination of \n<italic>n</i>\n-channel FeFETs (\n<italic>n</i>\n-FeFETs) and \n<italic>p</i>\n-channel FeFETs (\n<italic>p</i>\n-FeFETs). We first confirm that the reservoir computing performance of a single \n<italic>p</i>\n-FeFET is similar to that of a single \n<italic>n</i>\n-FeFET despite the smaller memory window. We demonstrate that combining both the reservoir states of \n<italic>n</i>\n- and \n<italic>p</i>\n-FeFETs enhances the performance of reservoir computing owing to the complementary behaviors of the two FeFETs: the \n<italic>p</i>\n-FeFET is turned ON when the \n<italic>n</i>\n-FeFET is turned OFF and vice versa, providing large current output for any gate voltage input. These complementary characteristics provide the reservoir states that exhibit clear nonlinear transformation of any given input. It is also found that additionally combining with techniques using inverted digital inputs can further enhance the performance.","PeriodicalId":13198,"journal":{"name":"IEEE Electron Device Letters","volume":null,"pages":null},"PeriodicalIF":4.1000,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Electron Device Letters","FirstCategoryId":"5","ListUrlMain":"https://ieeexplore.ieee.org/document/10614216/","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
Abstract
We investigate the potential of physical reservoir computing utilizing a combination of
n
-channel FeFETs (
n
-FeFETs) and
p
-channel FeFETs (
p
-FeFETs). We first confirm that the reservoir computing performance of a single
p
-FeFET is similar to that of a single
n
-FeFET despite the smaller memory window. We demonstrate that combining both the reservoir states of
n
- and
p
-FeFETs enhances the performance of reservoir computing owing to the complementary behaviors of the two FeFETs: the
p
-FeFET is turned ON when the
n
-FeFET is turned OFF and vice versa, providing large current output for any gate voltage input. These complementary characteristics provide the reservoir states that exhibit clear nonlinear transformation of any given input. It is also found that additionally combining with techniques using inverted digital inputs can further enhance the performance.
我们研究了利用 n 沟道场效应晶体管(n-FeFET)和 p 沟道场效应晶体管(p-FeFET)组合进行物理水库计算的潜力。我们首先证实,尽管内存窗口较小,但单个 p 型场效应晶体管的水库计算性能与单个 n 型场效应晶体管类似。我们证明,将 n 型场效应晶体管和 p 型场效应晶体管的贮存状态结合在一起可以提高贮存计算性能,这是因为这两种场效应晶体管具有互补性:当 n 型场效应晶体管关闭时,p 型场效应晶体管接通,反之亦然,从而为任何栅极电压输入提供大电流输出。这些互补特性提供了储能状态,可对任何给定输入进行明显的非线性转换。研究还发现,与使用反相数字输入的技术相结合可进一步提高性能。
期刊介绍:
IEEE Electron Device Letters publishes original and significant contributions relating to the theory, modeling, design, performance and reliability of electron and ion integrated circuit devices and interconnects, involving insulators, metals, organic materials, micro-plasmas, semiconductors, quantum-effect structures, vacuum devices, and emerging materials with applications in bioelectronics, biomedical electronics, computation, communications, displays, microelectromechanics, imaging, micro-actuators, nanoelectronics, optoelectronics, photovoltaics, power ICs and micro-sensors.