中国物理B ›› 2020, Vol. 29 ›› Issue (7): 78504-078504.doi: 10.1088/1674-1056/ab90e7

所属专题: SPECIAL TOPIC — Physics in neuromorphic devices

• SPECIAL TOPIC—Ultracold atom and its application in precision measurement • 上一篇    下一篇

In-memory computing to break the memory wall

Xiaohe Huang(黄晓合), Chunsen Liu(刘春森), Yu-Gang Jiang(姜育刚), Peng Zhou(周鹏)   

  1. 1 State Key Laboratory of ASIC and System, School of Microelectronics, Fudan University, Shanghai 200433, China;
    2 School of Computer Science, Fudan University, Shanghai 200433, China
  • 收稿日期:2020-04-05 修回日期:2020-05-02 出版日期:2020-07-05 发布日期:2020-07-05
  • 通讯作者: Peng Zhou E-mail:pengzhou@fudan.edu.cn
  • 基金资助:
    Project supported by the National Natural Science Foundation of China (Grant Nos. 61925402 and 61851402 ), Science and Technology Commission of Shanghai Municipality, China (Grant No. 19JC1416600), the National Key Research and Development Program of China (Grant No. 2017YFB0405600), and Shanghai Education Development Foundation and Shanghai Municipal Education Commission Shuguang Program, China (Grant No. 18SG01).

In-memory computing to break the memory wall

Xiaohe Huang(黄晓合)1, Chunsen Liu(刘春森)1,2, Yu-Gang Jiang(姜育刚)2, Peng Zhou(周鹏)1   

  1. 1 State Key Laboratory of ASIC and System, School of Microelectronics, Fudan University, Shanghai 200433, China;
    2 School of Computer Science, Fudan University, Shanghai 200433, China
  • Received:2020-04-05 Revised:2020-05-02 Online:2020-07-05 Published:2020-07-05
  • Contact: Peng Zhou E-mail:pengzhou@fudan.edu.cn
  • Supported by:
    Project supported by the National Natural Science Foundation of China (Grant Nos. 61925402 and 61851402 ), Science and Technology Commission of Shanghai Municipality, China (Grant No. 19JC1416600), the National Key Research and Development Program of China (Grant No. 2017YFB0405600), and Shanghai Education Development Foundation and Shanghai Municipal Education Commission Shuguang Program, China (Grant No. 18SG01).

摘要: Facing the computing demands of Internet of things (IoT) and artificial intelligence (AI), the cost induced by moving the data between the central processing unit (CPU) and memory is the key problem and a chip featured with flexible structural unit, ultra-low power consumption, and huge parallelism will be needed. In-memory computing, a non-von Neumann architecture fusing memory units and computing units, can eliminate the data transfer time and energy consumption while performing massive parallel computations. Prototype in-memory computing schemes modified from different memory technologies have shown orders of magnitude improvement in computing efficiency, making it be regarded as the ultimate computing paradigm. Here we review the state-of-the-art memory device technologies potential for in-memory computing, summarize their versatile applications in neural network, stochastic generation, and hybrid precision digital computing, with promising solutions for unprecedented computing tasks, and also discuss the challenges of stability and integration for general in-memory computing.

关键词: in-memory computing, non-volatile memory, device technologies, crossbar array

Abstract: Facing the computing demands of Internet of things (IoT) and artificial intelligence (AI), the cost induced by moving the data between the central processing unit (CPU) and memory is the key problem and a chip featured with flexible structural unit, ultra-low power consumption, and huge parallelism will be needed. In-memory computing, a non-von Neumann architecture fusing memory units and computing units, can eliminate the data transfer time and energy consumption while performing massive parallel computations. Prototype in-memory computing schemes modified from different memory technologies have shown orders of magnitude improvement in computing efficiency, making it be regarded as the ultimate computing paradigm. Here we review the state-of-the-art memory device technologies potential for in-memory computing, summarize their versatile applications in neural network, stochastic generation, and hybrid precision digital computing, with promising solutions for unprecedented computing tasks, and also discuss the challenges of stability and integration for general in-memory computing.

Key words: in-memory computing, non-volatile memory, device technologies, crossbar array

中图分类号:  (Superconducting logic elements and memory devices; microelectronic circuits)

  • 85.25.Hv
51.50.+v (Electrical properties)