CIF: Small: Realizing Chip-scale Bio-inspired Spiking Neural Networks with Monolithically Integrated Nano-scale Memristors

CIF：小型：利用单片集成纳米级忆阻器实现芯片级仿生尖峰神经网络

• 批准号: 1320987
• 负责人: Elisa Barney Smith
• 金额: $ 50万
• 依托单位: Boise State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2013
• 资助国家: 美国
• 起止时间: 2013-07-01 至 2017-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1320987&HistoricalAwards=false
• 关键词: CIF; Small; Realizing; Chip; scale

Computer processing power is so integrated into our daily lives that we hardly notice how it enables everything from a routine text message to solving "large data" problems that mandate immense supercomputing resources. And yet, the processing power of today's computers pales in comparison to that most advanced processor - the human brain. It is now believed that technological progress would enable us to take up the challenge of developing a new kind of computing architecture that functions more like the brain. In recent years, scientists have examined the electrical interaction between brain synapses and how biological neurons interact. They have also derived mathematical models to explain how these processes work. By employing these models in combination with a new device technology that exhibits similar electrical response to the neural synapses, this project will design entirely new computer processing chips that mimic how the brain processes information.We envision these chips performing pattern recognition with machine complexity an order of magnitude higher than traditional computing and digital signal processor (DSP) architectures. And even though the chip physical size and weight will match current processors, their power consumption will be orders of magnitude lower than with the von Neumann computing architecture that forms the basis for most of today's computer processors. Therefore, by mimicking the brain's billions of interconnections and pattern recognition capabilities, we may ultimately introduce a new paradigm in speed and power, and potentially enable systems that include the ability to learn, adapt, and respond to their environment.

计算机处理能力与我们的日常生活如此紧密地结合在一起，以至于我们几乎没有注意到它是如何实现从常规短信到解决需要巨大超级计算资源的“大数据”问题的一切的。然而，与最先进的处理器--人脑相比，今天计算机的处理能力相形见绌。现在人们相信，技术进步将使我们能够接受开发一种新的计算架构的挑战，这种架构的功能更像大脑。近年来，科学家们研究了大脑突触之间的电相互作用以及生物神经元如何相互作用。他们还推导出数学模型来解释这些过程是如何工作的。通过将这些模型与对神经突触表现出类似电响应的新设备技术相结合，该项目将设计出全新的计算机处理芯片，模拟大脑如何处理信息。我们设想这些芯片执行模式识别，其机器复杂度比传统计算和数字信号处理器（DSP）架构高一个数量级。尽管芯片的物理尺寸和重量将与当前的处理器相匹配，但它们的功耗将比冯·诺依曼计算架构低几个数量级，冯·诺依曼计算架构是当今大多数计算机处理器的基础。因此，通过模仿大脑数十亿的互连和模式识别能力，我们最终可能会在速度和功率方面引入一种新的范式，并可能使系统能够学习，适应和响应环境。