Eager: High-Density Electronic Synapse Arrays Using Nanoscale Devices

Eager：使用纳米级器件的高密度电子突触阵列

• 批准号: 0950305
• 负责人: H S Philip Wong
• 金额: $ 5.56万
• 依托单位: Stanford University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-09-01 至 2010-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0950305&HistoricalAwards=false
• 关键词: Eager; Density; Electronic; Synapse; Arrays

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).This EArly-concept Grants for Exploratory Research (EAGER) award seeks to use nanoscale electronics technology to develop brain-like computing structures that possess the density and plasticity that allow arrays of synapses to perform sensing, pattern recognition, and motor control at performance levels beyond what is possible with traditional computers. The goal is to determine if nanoelectronics configured in a self-assembled crossbar array can achieve the density and plasticity of synapses. To address plasticity, the research focuses on phase-change and metal-oxide materials for neuromorphic systems. To address density, the project investigates ultra-dense interconnects for self-assembled nanoelectronics.With respect to intellectual merit, the research builds on recent progress in the use of phase-change and metal-oxide memory cells for nonvolatile memory, but takes the technology in a different direction to use phase-change and metal-oxide memory cells for implementing electronic synapse arrays for brain-like computing. The research uses an experimental approach.With respect to broader impact, this research has the potential to dramatically alter how computation is done for a number of tasks, such as pattern recognition and motor control. An undergraduate student from an underrepresented group will be engaged in the research. The project also facilitates multidisciplinary interaction between nanoelectronics and bioengineering researchers.

该奖项根据 2009 年《美国复苏和再投资法案》（公法 111-5）提供资助。这项早期概念探索性研究资助 (EAGER) 奖项旨在利用纳米级电子技术开发类脑计算结构，该结构具有密度和可塑性，允许突触阵列以超越传统计算机的性能水平执行传感、模式识别和运动控制。 目标是确定自组装交叉阵列中配置的纳米电子学是否可以实现突触的密度和可塑性。为了解决可塑性问题，该研究重点关注神经形态系统的相变和金属氧化物材料。 为了解决密度问题，该项目研究了自组装纳米电子学的超密集互连。就智力价值而言，该研究建立在使用相变和金属氧化物存储单元用于非易失性存储器的最新进展的基础上，但将该技术引向了不同的方向，即使用相变和金属氧化物存储单元来实现用于类脑计算的电子突触阵列。 该研究采用实验方法。就更广泛的影响而言，这项研究有可能极大地改变许多任务的计算方式，例如模式识别和运动控制。 来自代表性不足群体的一名本科生将参与这项研究。 该项目还促进纳米电子学和生物工程研究人员之间的多学科互动。