{"title":"Monolithic Three-Dimensional Imaging System: Carbon Nanotube Computing Circuitry Integrated Directly Over Silicon Imager","authors":"T. Srimani, G. Hills, C. Lau, M. Shulaker","doi":"10.23919/VLSIT.2019.8776514","DOIUrl":null,"url":null,"abstract":"Here we show a hardware prototype of a monolithic three-dimensional (3D) imaging system that integrates computing layers directly in the back-end-of-line (BEOL) of a conventional silicon imager. Such systems can transform imager output from raw pixel data to highly processed information. To realize our imager, we fabricate 3 vertical circuit layers directly on top of each other: a bottom layer of silicon pixels followed by two layers of CMOS carbon nanotube FETs (CNFETs) (comprising 2,784 CNFETs) that perform in-situ edge detection in real-time, before storing data in memory. This approach promises to enable image classification systems with improved nrocessing latencies.","PeriodicalId":6752,"journal":{"name":"2019 Symposium on VLSI Technology","volume":"33 1","pages":"T24-T25"},"PeriodicalIF":0.0000,"publicationDate":"2019-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"12","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 Symposium on VLSI Technology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.23919/VLSIT.2019.8776514","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 12
Abstract
Here we show a hardware prototype of a monolithic three-dimensional (3D) imaging system that integrates computing layers directly in the back-end-of-line (BEOL) of a conventional silicon imager. Such systems can transform imager output from raw pixel data to highly processed information. To realize our imager, we fabricate 3 vertical circuit layers directly on top of each other: a bottom layer of silicon pixels followed by two layers of CMOS carbon nanotube FETs (CNFETs) (comprising 2,784 CNFETs) that perform in-situ edge detection in real-time, before storing data in memory. This approach promises to enable image classification systems with improved nrocessing latencies.