NIRT: Devices and Architectures for Neuromorphic Circuits with Nanoelectronic Components

具有纳米电子元件的神经形态电路的器件和体系结构本提案是对NSF 03-043 NIRT类纳米科学与工程倡议的响应。初步研究表明，混合半导体/分子(CMOL)电路可以作为神经形态网络(CrossNets)的基础，能够以前所未有的密度(超过1012个有源器件/cm2)和以可接受的功耗(低于100W/cm2)的性能(高达1020次/cm2的基本运算)进行先进的混合信号信息处理。该项目解决了CMOL CrossNet开发的两个关键问题：(I)在预制金属纳米线结构上自组装的单分子锁存开关的实验研究，以及这些组件的传输特性的表征；以及(Ii)数字和混合信号CMOL电路的先进架构的开发，包括数字处理器和混合信号神经形态网络。预计在未来十年的某个时候，硅数字技术的指数级进步将撞上基本物理和技术限制的“红砖墙”。该项目将探索一个可能的方向，通过从纯半导体数字芯片向混合半导体/分子集成电路转移，将增长扩展到纳米级设备。在这样的电路中，最简单、数量最多的器件将被化学合成，然后从溶液中在带有预制纳米线织物的硅芯片上自组装。该项目解决了这一新的令人兴奋的领域的两个关键问题：分子器件的合成和自组装，以及为未来的混合电路开发创新的、受生物启发的架构。

Devices and Architectures for Neuromorphic Circuits with Nanoelectronic ComponentsThis proposal was received in response to Nanoscale Science and Engineering initiative, NSF 03-043, category NIRT. Preliminary studies indicate that hybrid semiconductor/molecular ("CMOL") circuits may serve as a basis for neuromorphic networks ("CrossNets") capable of advanced mixed-signal information processing with unprecedented density (beyond 1012 active devices per cm2) and performance (up to 1020 elementary operations per second per cm2) at acceptable power consumption (below 100 W/cm2). The project addresses two key issues of the CMOL CrossNet development: (i) experimental study of single-molecule latching switches self-assembled on prefabricated metallic nanowire structures, and characterization of transport properties of these components, and (ii) development of advanced architectures for digital and mixed-signal CMOL circuits, including digital processors and mixed-signal neuromorphic networks. The exponential progress of silicon digital technology is expected to run, some time during the next decade, into the "red brick wall" of fundamental physical and technical limitations. The project will explore one possible direction to extend the growth to nanoscale devices, by transferring from purely semiconductor digital chips to hybrid semiconductor/molecular integrated circuits. In such circuits the simplest and most numerous devices would be chemically synthesized and then self-assemble, from solution, on silicon chips with prefabricated nanowire fabrics. This project addresses two key issues of this new exciting field: the synthesis and self-assembly of molecular devices and the development of innovative, bio-inspired architectures for future hybrid circuits.