NSF Convergence Accelerator Track M: Enabling novel photonic neuromorphic devices through bridging DNA-programmable assembly and nanofabrication

NSF 融合加速器轨道 M：通过桥接 DNA 可编程组装和纳米制造实现新型光子神经形态设备

• 批准号: 2344415
• 负责人: Oleg Gang
• 金额: $ 65万
• 依托单位: Columbia University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2024
• 资助国家: 美国
• 起止时间: 2024-01-15 至 2024-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2344415&HistoricalAwards=false
• 关键词: NSF; Convergence; Accelerator; Track; Enabling

Increasing demand for data processing cannot be satisfied by current technologies due to the incompatibility between new brain-inspired information-processing algorithms, hardware operating on traditional principles, and prohibitive energy requirements The project aims to address this impasse. Specifically, it aims to develop energy-efficient optical brain-inspired (neuromorphic) computing devices through the establishment of a 3D nanofabrication platform that combines recent advances in DNA-programmable assembly and conventional lithographic methods. The developed nanofabrication methodology will be applied to solve outstanding challenges in designing and realizing novel optical metamaterials and their device-level integration for neuromorphic computing devices. By integrating optically active nanoscale components and controlling 3D organization at different scales, from nanometers to millimeters, the proposed approach will offer unprecedented opportunities to create energy-efficient, parallel, fast, and secure neuromorphic computing devices for diverse computation-intensive tasks. The project also will offer training and professional development opportunities in STEM disciplines for undergraduate and graduate students from diverse backgrounds.Part 2. The proposed project seeks to increase optical computing density drastically by using novel neuromorphic computing devices. Such devices will be realized through specifically engineered 3D optically active nanostructured media. A neuromorphic computation, inspired by neuron information processing, is performed in this project by light propagating through the engineered media. The approach will enable a new computing paradigm suitable for several types of highly intense mathematical operations, including matrix multiplication, recurrent neural networks, and solving integral equations that can be applied broadly for image recognition. To realize such a computational approach, a DNA-assisted nanofabrication platform will be established. The platform will effectively introduce a required set of nanomaterials and methods in fabricating the designed optical neuromorphic processors. The project will establish methods for creating: (i) Integration of DNA programmable self-assembly and lithographic nanofabrication for fabricating arbitrarily defined surface-bound 3D nanostructures; (ii) Designed metamaterials with grayscale and complex optical refractive indices; (ii) Prescribed 3D nano- and mesoscale organizations of metamaterials over large areas (to a few millimeters) for realizing optical neuromorphic computing devices.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

目前的技术无法满足日益增长的数据处理需求，因为受大脑启发的新信息处理算法、按传统原理运行的硬件和令人望而却步的能源需求之间不兼容，该项目旨在解决这一僵局。具体地说，它的目标是通过建立一个结合了DNA可编程组装和传统光刻方法的最新进展的3D纳米制造平台，开发高能效的光学大脑启发(神经形态)计算设备。开发的纳米制造方法将用于解决为神经形态计算设备设计和实现新型光学超材料及其器件级集成方面的突出挑战。通过集成光学活性纳米级组件并控制从纳米到毫米的不同尺度的3D组织，所提出的方法将提供前所未有的机会，为各种计算密集型任务创造节能、并行、快速和安全的神经形态计算设备。该项目还将为来自不同背景的本科生和研究生提供STEM学科的培训和专业发展机会。第二部分：拟议的项目旨在通过使用新的神经形态计算设备大幅提高光计算密度。这种装置将通过专门设计的3D光学活性纳米结构介质来实现。在这个项目中，受神经元信息处理的启发，通过光在工程介质中传播来执行神经形态计算。这种方法将实现一种新的计算范例，适用于几种类型的高强度数学运算，包括矩阵乘法、递归神经网络和求解可广泛应用于图像识别的积分方程组。为了实现这种计算方法，将建立一个DNA辅助纳米制造平台。该平台将有效地引入一套所需的纳米材料和方法来制造所设计的光学神经形态处理器。该项目将建立以下方法：(I)集成DNA可编程自组装和光刻纳米制造，以制造任意定义的表面结合的3D纳米结构；(Ii)设计具有灰度级和复杂光学折射率的超材料；(Ii)用于实现光学神经形态计算设备的规定的3D纳米和中尺度超材料组织。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。