基于SRAM的感存算一体化技术综述

杨兴华; 杨子翼; 苏海津; 姜炜煌; 张静; 魏琦; 骆丽; 王忠静; 吕华芳; 乔飞

doi:10.11999/JEIT220815

基于SRAM的感存算一体化技术综述

doi: 10.11999/JEIT220815

杨兴华¹,
杨子翼^2, ,,
苏海津³,
姜炜煌⁴,
张静²,
魏琦⁴,
骆丽³,
王忠静⁴,
吕华芳⁴,
乔飞⁴

1.
北京林业大学理学院北京 100091
2.
北方工业大学信息学院北京 100041
3.
北京交通大学电子信息工程学院北京 100044
4.
清华大学北京 100084

基金项目: 国家自然科学基金(92164203)，清华大学-宁夏银川水联网数字治水联合研究院基金项目(SKL-IOW-2020TC2003)

详细信息

作者简介:
杨兴华：男，讲师，研究方向为近似计算电路系统设计

张静：女，教授，研究方向为智能传感技术、第三代半导体器件

魏琦：男，副研究员，研究方向为集成电路设计

骆丽：女，教授，研究方向为微电子学与固体电子学

王忠静：男，教授，研究方向为水利物联网感知

吕华芳：女，高级工程师，研究方向为水利物联网感知

乔飞：男，副研究员，研究方向为智能感知集成电路与系统

通讯作者:
杨子翼　yangziyi0128@gmail.com

中图分类号: TN403
计量
- 文章访问数: 1614
- HTML全文浏览量: 1125
- PDF下载量: 403
- 被引次数: 0
出版历程
- 收稿日期: 2022-06-21
- 修回日期: 2022-12-10
- 录用日期: 2022-12-20
- 网络出版日期: 2022-12-23
- 刊出日期: 2023-08-21

Review of the Fused Technology of Sensing, Storage and Computing Based on SRAM

1.
College of Science, Beijing Forestry University, Beijing 100091, China
2.
School of Information, North China University of Technology, Beijing 100041, China
3.
School of Electronic and Information Engineering, Beijing Jiaotong University, Beijing 100044, China
4.
Tsinghua University, Beijing 100084, China

Funds: The National Natural Science Foundation of China (92164203)，The Tsinghua University-Ningxia Yinchuan Water Network Digital Water Control Joint Research Institute Fund Project (SKL-IOW-2020TC2003)

摘要

摘要: 基于SRAM(静态随机存取)存储器的感存算一体化芯片架构将传感、存储和计算功能结合，通过使存储单元具备计算能力，避免了计算过程中数据的搬移，解决了冯诺依曼架构所面临的“存储墙”的问题。该结构与传感器部分结合，可以实现超高速、超低功耗的运算能力。SRAM存储器相较于其他存储器在速度方面具有较大优势，主要体现在该架构能够实现较高的能效比，在精度增强后可以保证较高精度，适用于低功耗高性能要求下的大算力场景设计。该文调研了近几年来关于感存算一体化的研究，介绍了传统感知系统和持续感知系统及感算共融系统，并介绍了基于SRAM存储器的感存算一体芯片最常见的几种计算单元结构，在电压域、电荷域和数字域考察了基于SRAM的感存算一体的研究发展，进行分析对比其优劣势，结合调研分析讨论了该领域的未来发展方向。
- 感存算一体 /
- SRAM存储器 /
- 冯诺依曼计算架构
Abstract: The integrated chip architecture based on SRAM memory combines sensing, storage and computing functions, solving the problem of “storage wall” faced by the Von Neumann architecture by enabling the memory unit to have computing power and avoiding the transfer of data during the calculation process. This structure is combined with the sensor part to achieve ultra-high-speed, ultra-low-power computing power. SRAM memory has great advantages in terms of speed compared to other memory, mainly reflected in the fact that the architecture can achieve a high energy efficiency ratio, which can ensure high accuracy after the accuracy is enhanced, and is suitable for large computing power scenario design under the requirements of low power consumption and high performance. This paper introduces the structure of the sensor-memory-computing chip based on SRAM memory, investigates the research and development of SRAM-based sensor-memory-computing integration in the voltage domain, charge domain and digital domain, and discusses the future development direction of this field.
- Integration of sensing, storage and computing /
- SRAM Memory /
- Von Neumann

HTML全文

图 1 传统感知系统的处理流程

下载: 全尺寸图片幻灯片

图 2 持续感知系统与传统感知系统的功耗模式对

下载: 全尺寸图片幻灯片

图 3 SRAM 6T单元电路

下载: 全尺寸图片幻灯片

图 4 分裂式6T SRAM单元结构

下载: 全尺寸图片幻灯片

图 5 4+2T 存储器单元结构

下载: 全尺寸图片幻灯片

图 6 双7T SRAM 单元

下载: 全尺寸图片幻灯片

图 7 8T SRAM单元电路

下载: 全尺寸图片幻灯片

图 8 SRAM 9T单元电路

下载: 全尺寸图片幻灯片

图 9 开关电容阵列实现电荷域计算

下载: 全尺寸图片幻灯片

图 10 电荷域计算10T单元

下载: 全尺寸图片幻灯片

图 11 电荷域计算10T单元

下载: 全尺寸图片幻灯片

图 12 串行计算的可重构数字存内计算体系结构

下载: 全尺寸图片幻灯片

参考文献(40)

[1]	HAENSCH W, GOKMEN T, and PURI R. The next generation of deep learning hardware: Analog computing[J]. Proceedings of the IEEE, 2019, 107(1): 108–122. doi: 10.1109/JPROC.2018.2871057
[2]	MA Yufei, DU Yuan, DU Li, et al. In-memory computing: The next-generation AI computing paradigm[C]. 2020 on Great Lakes Symposium on VLSI, New York, USA, 2020: 265–270.
[3]	李桂宏, 乔飞. 面向边缘智能设备的持续感知集成电路与系统[J]. 微纳电子与智能制造, 2019, 1(2): 47–61. doi: 10.19816/j.cnki.10-1594/tn.2019.02.007 LI Guihong and QIAO Fei. Continuous perception integrated circuits and systems for edge-computing smart devices[J]. Micro/Nano Electronics and Intelligent Manufacturing, 2019, 1(2): 47–61. doi: 10.19816/j.cnki.10-1594/tn.2019.02.007
[4]	CHOI J. Review of low power image sensors for always-on imaging[C]. 2016 International SoC Design Conference, Jeju, Korea (South), 2016.
[5]	PAUL S, HONKOTE V, KIM R G, et al. A sub-cm³ energy-harvesting stacked wireless sensor node featuring a near-threshold voltage IA-32 microcontroller in 14-nm tri-gate CMOS for always-on always-sensing applications[J]. IEEE Journal of Solid-State Circuits, 2017, 52(4): 961–971. doi: 10.1109/JSSC.2016.2638465
[6]	CHOI J, SHIN J, KANG Dongwu, et al. Always-on CMOS image sensor for mobile and wearable devices[J]. IEEE Journal of Solid-State Circuits, 2016, 51(1): 130–140. doi: 10.1109/JSSC.2015.2470526
[7]	LUO Yi and MIRABBASI S. Always-on CMOS image sensor pixel design for pixel-wise binary coded exposure[C]. 2017 IEEE International Symposium on Circuits and Systems, Baltimore, USA, 2017: 1–4.
[8]	YOUNG C, OMID-ZOHOOR A, LAJEVARDI P, et al. 5.3 a data-compressive 1.5b/2.75b log-gradient QVGA image sensor with multi-scale readout for always-on object detection[C]. 2019 IEEE International Solid-State Circuits Conference, San Francisco, USA, 2019: 98–100.
[9]	SHI Weisong, CAO Jie, ZHANG Quan, et al. Edge computing: Vision and challenges[J]. IEEE Internet of Things Journal, 2016, 3(5): 637–646. doi: 10.1109/JIOT.2016.2579198
[10]	CHIANG M and ZHANG Tao. Fog and IoT: An overview of research opportunities[J]. IEEE Internet of Things Journal, 2016, 3(6): 854–864. doi: 10.1109/JIOT.2016.2584538
[11]	ZHU Bowen, WANG Hong, LIU Yaqing, et al. Skin-inspired haptic memory arrays with an electrically reconfigurable architecture[J]. Advanced Materials, 2016, 28(8): 1559–1566. doi: 10.1002/adma.201504754
[12]	JIANG Chengming, LI Qikun, SUN Nan, et al. A high-performance bionic pressure memory device based on piezo-OLED and piezo-memristor as luminescence-fish neuromorphic tactile system[J]. Nano Energy, 2020, 77: 105120. doi: 10.1016/j.nanoen.2020.105120
[13]	SUN Yihui, ZHENG Xin, YAN Xiaoqin, et al. Bioinspired tribotronic resistive switching memory for self-powered memorizing mechanical stimuli[J]. ACS Applied Materials & Interfaces, 2017, 9(50): 43822–43829. doi: 10.1021/acsami.7b15269
[14]	WAN Changjin, CAI Pingqiang, GUO Xintong, et al. An artificial sensory neuron with visual-haptic fusion[J]. Nature Communications, 2020, 11(1): 4602. doi: 10.1038/s41467-020-18375-y
[15]	ZHOU Feichi, ZHOU Zheng, CHEN Jiewei, et al. Optoelectronic resistive random access memory for neuromorphic vision sensors[J]. Nature Nanotechnology, 2019, 14(8): 776–782. doi: 10.1038/s41565-019-0501-3
[16]	LORENZI P, SUCRE V, ROMANO G, et al. Memristor based neuromorphic circuit for visual pattern recognition[C]. 2015 International Conference on Memristive Systems, Paphos, Cyprus, 2015: 1–2.
[17]	HALAWANI Y, MOHAMMAD B, AL-QUTAYRI M, et al. Memristor-based hardware accelerator for image compression[J]. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2018, 26(12): 2749–2758. doi: 10.1109/TVLSI.2018.2835572
[18]	WANG Wei, PEDRETTI G, MILO V, et al. Learning of spatiotemporal patterns in a spiking neural network with resistive switching synapses[J]. Science Advances, 2018, 4(9): eaat4752. doi: 10.1126/sciadv.aat4752
[19]	LI Qin, YANG Yuze, LAN Tianxiang, et al. MSP-MFCC: Energy-efficient MFCC feature extraction method with mixed-signal processing architecture for wearable speech recognition applications[J]. IEEE Access, 2020, 8: 48720–48730. doi: 10.1109/ACCESS.2020.2979799
[20]	CHANG Mengfan, CHEN C F, CHANG Tinghao, et al. 17.3 A 28nm 256kb 6T-SRAM with 280mV improvement in VMIN using a dual-split-control assist scheme[C]. 2015 IEEE International Solid-State Circuits Conference–(ISSCC) Digest of Technical Papers, San Francisco, USA, 2015: 1–3.
[21]	KHWA W S, CHEN Jiajing, LI Jiafang, et al. A 65nm 4Kb algorithm-dependent computing-in-memory SRAM unit-macro with 2.3ns and 55.8TOPS/W fully parallel product-sum operation for binary DNN edge processors[C]. 2018 IEEE International Solid–State Circuits Conference, San Francisco, USA, 2018: 496–498.
[22]	SI Xin, KHWA W S, CHEN Jiajing, et al. A dual-split 6T SRAM-based computing-in-memory unit-macro with fully parallel product-sum operation for binarized DNN edge processors[J]. IEEE Transactions on Circuits and Systems I:Regular Papers, 2019, 66(11): 4172–4185. doi: 10.1109/TCSI.2019.2928043
[23]	DONG Qing, JELOKA S, SALIGANE M, et al. A 0.3V VDDmin 4+2T SRAM for searching and in-memory computing using 55nm DDC technology[C]. 2017 Symposium on VLSI Circuits, Kyoto, Japan, 2017: C160–C161.
[24]	NAJAFI D and EBRAHIMI B. A low-leakage 6T SRAM cell for in-memory computing with high stability[C]. The 2021 29th Iranian Conference on Electrical Engineering, Tehran, Iran, 2021: 98–102.
[25]	YU Chengshuo, YOO T, CHAI K T C, et al. A 65-nm 8T SRAM compute-in-memory macro with column ADCs for processing neural networks[J]. IEEE Journal of Solid-State Circuits, 2022, 57(11): 3466–3476. doi: 10.1109/JSSC.2022.3162602
[26]	YU Chengshuo, CHAI K T C, KIM T T H, et al. A zero-skipping reconfigurable SRAM in-memory computing macro with binary-searching ADC[C]. The IEEE 47th European Solid State Circuits Conference, Grenoble, France, 2021: 131–134.
[27]	GUPTA A K and ACHARYA A. Exploration of 9T SRAM cell for in memory computing application[C]. 2021 Devices for Integrated Circuit, Kalyani, India, 2021: 461–465.
[28]	DONG Qing, SINANGIL M E, ERBAGCI B, et al. 15.3 A 351TOPS/W and 372.4GOPS compute-in-memory SRAM macro in 7nm FinFET CMOS for machine-learning applications[C]. 2020 IEEE International Solid- State Circuits Conference, San Francisco, USA, 2020: 242–244.
[29]	JIA Hongyang, OZATAY M, TANG Yinqi, et al. 15.1 A programmable neural-network inference accelerator based on scalable in-memory computing[C]. 2021 IEEE International Solid- State Circuits Conference, San Francisco, USA, 2021: 236–238.
[30]	KIM J and PARK J. A charge-domain 10T SRAM based in-memory-computing macro for low energy and highly accurate DNN inference[C]. The 2021 18th International SoC Design Conference, Jeju Island, Korea (South), 2021: 89–90.
[31]	OH H, KIM H, AHN D, et al. Energy-efficient in-memory binary neural network accelerator design based on 8T2C SRAM cell[J]. IEEE Solid-State Circuits Letters, 2022, 2,5: 70–73. doi: 10.1109/LSSC.2022.3161592
[32]	KIM H, CHEN Qian, YOO T, et al. A 1–16b precision reconfigurable digital in-memory computing macro featuring column-MAC architecture and bit-serial computation[C]. The IEEE 45th European Solid State Circuits Conference, Cracow, Poland, 2019: 345–348.
[33]	CHIH Y D, LEE P H, FUJIWARA H, et al. 16.4 An 89TOPS/W and 16.3TOPS/mm² all-digital SRAM-based full-precision compute-in memory macro in 22nm for machine-learning edge applications[C]. 2021 IEEE International Solid- State Circuits Conference, San Francisco, USA, 2021: 252–254.
[34]	KIM H, YOO T, KIM T T H, et al. Colonnade: A reconfigurable SRAM-based digital bit-serial compute-in-memory macro for processing neural networks[J]. IEEE Journal of Solid-State Circuits, 2021, 56(7): 2221–2233. doi: 10.1109/JSSC.2021.3061508
[35]	LIN Zhiting, ZHU Zhiyong, ZHAN Honglan, et al. Two-direction in-memory computing based on 10T SRAM with horizontal and vertical decoupled read ports[J]. IEEE Journal of Solid-State Circuits, 2021, 56(9): 2832–2844. doi: 10.1109/JSSC.2021.3061260
[36]	FUJIWARA H, MORI H, ZHAO Weichang, et al. A 5-nm 254-TOPS/W 221-TOPS/mm² fully-digital computing-in-memory macro supporting wide-range dynamic-voltage-frequency scaling and simultaneous MAC and write operations[C]. 2022 IEEE International Solid- State Circuits Conference, San Francisco, USA, 2022: 1–3.
[37]	YAN Bonan, HSU J L, YU Pangcheng, et al. A 1.041-Mb/mm2 27.38-TOPS/W signed-INT8 dynamic-logic-based ADC-less SRAM compute-in-memory macro in 28nm with reconfigurable bitwise operation for AI and embedded applications[C]. 2022 IEEE International Solid- State Circuits Conference, San Francisco, USA, 2022: 188–190.
[38]	龚龙庆, 徐伟栋, 娄冕. SRAM存内计算技术综述[J]. 微电子学与计算机, 2021, 38(9): 1–7. doi: 10.19304/j.cnki.issn1000-7180.2021.09.001 GONG Longqing, XU Weidong, and LOU Mian. An overview of SRAM in-memory computing[J]. Microelectronics &Computer, 2021, 38(9): 1–7. doi: 10.19304/j.cnki.issn1000-7180.2021.09.001
[39]	周正, 丛瑛瑛. 存内计算技术发展趋势分析[J]. 信息通信技术与政策, 2019(9): 65–68. doi: 10.3969/j.issn.1008-9217.2019.09.016 ZHOU Zheng and CONG Yingying. Analysis on the development trend of Computing in-memory[J]. Information and Communications Technology and Policy, 2019(9): 65–68. doi: 10.3969/j.issn.1008-9217.2019.09.016
[40]	张章, 李超, 韩婷婷, 等. 基于忆阻器的感存算一体技术综述[J]. 电子与信息学报, 2021, 43(6): 1498–1509. doi: 10.11999/JEIT201102 ZHANG Zhang, LI Chao, HAN Tingting, et al. Review of the fused technology of sensing, storage and computing based on memristor[J]. Journal of Electronics &Information Technology, 2021, 43(6): 1498–1509. doi: 10.11999/JEIT201102