CAREER: Efficient, Dynamic, Robust, and On-Device Continual Deep Learning with Non-Volatile Memory based In-Memory Computing System

职业：使用基于非易失性内存的内存计算系统进行高效、动态、鲁棒、设备上持续深度学习

• 批准号: 2144751
• 负责人: Deliang Fan
• 金额: $ 50万
• 依托单位: Arizona State University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-01-15 至 2023-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2144751&HistoricalAwards=false
• 关键词: CAREER; Efficient; Dynamic; Robust; Device

This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2).Over past decades, there have existed grand challenges in developing high performance and energy-efficient computing solutions for big-data processing. Meanwhile, owing to the boom in artificial intelligence (AI), especially Deep Neural Networks (DNNs), such big-data processing requires efficient, intelligent, fast, dynamic, robust, and on-device adaptive cognitive computing. However, those requirements are not sufficiently satisfied by existing computing solutions due to the well-known power wall in silicon-based semiconductor devices, the memory wall in traditional Von-Neuman computing architectures, and computation-/memory-intensive DNN computing algorithms. This project aims to foster a systematic breakthrough in developing AI-in-Memory computing systems, through collaboratively developing ahybrid in-memory computing (IMC) hardware platform integrating the benefits of emerging non-volatile resistive memory (RRAM) and Static Random Access Memory (SRAM) technologies, as well as incorporating IMC-aware deep-learning algorithm innovations. The overarching goal of this project is to design, implement, and experimentally validate a new hybrid in-memory computing system that is collaboratively optimized for energy efficiency, inference accuracy, spatiotemporal dynamics, robustness, and on-device learning, which will greatly advance AI-based big-data processing fields such as computer vision, autonomous driving, robotics, etc. The research will also be extended into an educational platform, providing a user-friendly learning framework, and will serve the educational objectives for K-12 students, undergraduate, graduate, and under-represented students.This project will advance knowledge and produce scientific principles and tools for a new paradigm of AI-in-Memory computing featuring significant improvements in energy efficiency, speed, dynamics, robustness, and on-device learning capability. This cross-layer project spans from device, circuit, and architecture to DNN algorithm exploration. First, a hybrid RRAM-SRAM based in-memory computing chip will be designed, optimized, and fabricated. Second, based on this new computing platform, the on-device spatiotemporal dynamic neural network structure will be developed to provide an enhanced run-time computing profile (latency, resource allocation, working load, power budget, etc.), as well as improve the robustness of the system against hardware intrinsic and adversarial noise injection. Then, efficient on-device learning methodologies with the developed computing platform will be investigated. In the last thrust, an end-to-end DNN training, optimization, mapping, and evaluation CAD tool will be developed that integrates the developed hardware platform and algorithm innovations, for optimizing the software and hardware co-designs to achieve the user-defined multi-objectives in latency, energy efficiency, dynamics, accuracy, robustness, on-device adaption, etc.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.

该奖项的全部或部分资金来源于《2021 年美国救援计划法案》（公法 117-2）。过去几十年来，开发用于大数据处理的高性能和高能效计算解决方案面临着巨大的挑战。同时，由于人工智能（AI）特别是深度神经网络（DNN）的蓬勃发展，此类大数据处理需要高效、智能、快速、动态、鲁棒和设备端自适应认知计算。然而，由于众所周知的硅基半导体器件中的功率墙、传统冯诺依曼计算架构中的内存墙以及计算/内存密集型DNN计算算法，现有的计算解决方案并不能充分满足这些要求。该项目旨在通过合作开发融合新兴非易失性电阻存储器（RRAM）和静态随机存取存储器（RRAM）优点的混合内存计算（IMC）硬件平台，促进开发人工智能内存计算系统的系统性突破。 SRAM）技术，以及结合 IMC 感知的深度学习算法创新。该项目的总体目标是设计、实现和实验验证一种新型混合内存计算系统，该系统针对能源效率、推理准确性、时空动态、鲁棒性和设备上学习进行协作优化，这将极大地推进人工智能的发展基于大数据处理的计算机视觉、自动驾驶、机器人等领域。该研究还将延伸到教育平台，提供人性化的学习框架，服务于K-12学生的教育目标，本科生,该项目将推进知识进步，并为人工智能内存计算的新范式提供科学原理和工具，该范式在能源效率、速度、动态性、鲁棒性和设备学习能力方面都有显着改进。这个跨层项目涵盖了从器件、电路、架构到 DNN 算法的探索。首先，将设计、优化和制造基于混合 RRAM-SRAM 的内存计算芯片。其次，基于这个新的计算平台，将开发设备上的时空动态神经网络结构，以提供增强的运行时计算配置文件（延迟、资源分配、工作负载、功率预算等），并改进系统针对硬件固有和对抗性噪声注入的鲁棒性。然后，将研究具有开发的计算平台的高效设备上学习方法。最后，将开发端到端的 DNN 训练、优化、映射和评估 CAD 工具，集成已开发的硬件平台和算法创新，用于优化软硬件协同设计，以实现用户定义的目标。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。

项目成果

DSPIMM: A Fully Digital SParse In-Memory Matrix Vector Multiplier for Communication Applications

DSPIMM：用于通信应用的全数字稀疏内存矩阵向量乘法器

• 发表时间: 2023
• 期刊: 2023 60th ACM/IEEE Design Automation Conference (DAC
• 作者: Sridharan, Amitesh;Zhang, Fan;Sui, Yang;Yuan, Bo;Fan, Deliang
• 通讯作者: Fan, Deliang

DA3: Dynamic Additive Attention Adaption for Memory-Efficient On-Device Multi-Domain Learning

DA3：动态加性注意力适应，实现内存高效的设备上多域学习

• DOI: 10.1109/cvprw56347.2022.00295
• 发表时间: 2022
• 期刊: IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR
• 作者: Yang, Li' Rakin;Fan, Deliang
• 通讯作者: Fan, Deliang

A 1.23-GHz 16-kb Programmable and Generic Processing-in-SRAM Accelerator in 65nm

采用 65nm 工艺的 1.23GHz 16kb 可编程通用 SRAM 处理加速器

• DOI: 10.1109/esscirc55480.2022.9911440
• 发表时间: 2022
• 期刊: 2022- IEEE 48th European Solid State Circuits Conference (ESSCIRC
• 作者: Sridharan, Amitesh;Angizi, Shaahin;Cherupally, Sai Kiran;Zhang, Fan;Seo, Jae-Sun;Fan, Deliang
• 通讯作者: Fan, Deliang

XMA2: A crossbar-aware multi-task adaption framework via 2-tier masks

• DOI: 10.3389/felec.2022.1032485
• 发表时间: 2022-12
• 作者: Fan Zhang;Li Yang;Jian Meng;J.-s. Seo;Yu Cao;Deliang Fan

Hybrid RRAM/SRAM in-Memory Computing for Robust DNN Acceleration

• DOI: 10.1109/tcad.2022.3197516
• 发表时间: 2022-11
• 期刊: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
• 影响因子: 2.9
• 作者: Gokul Krishnan;Zhenyu Wang;Injune Yeo;Li Yang-;Jian Meng;Maximilian Liehr;R. Joshi;N. Cady;Deliang Fan;Jae-sun Seo;Yu Cao