1
|
Yang Y, Zhu F, Zhang X, Chen P, Wang Y, Zhu J, Ding Y, Cheng L, Li C, Jiang H, Wang Z, Lin P, Shi T, Wang M, Liu Q, Xu N, Liu M. Firing feature-driven neural circuits with scalable memristive neurons for robotic obstacle avoidance. Nat Commun 2024; 15:4318. [PMID: 38773067 PMCID: PMC11109161 DOI: 10.1038/s41467-024-48399-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/18/2023] [Accepted: 04/30/2024] [Indexed: 05/23/2024] Open
Abstract
Neural circuits with specific structures and diverse neuronal firing features are the foundation for supporting intelligent tasks in biology and are regarded as the driver for catalyzing next-generation artificial intelligence. Emulating neural circuits in hardware underpins engineering highly efficient neuromorphic chips, however, implementing a firing features-driven functional neural circuit is still an open question. In this work, inspired by avoidance neural circuits of crickets, we construct a spiking feature-driven sensorimotor control neural circuit consisting of three memristive Hodgkin-Huxley neurons. The ascending neurons exhibit mixed tonic spiking and bursting features, which are used for encoding sensing input. Additionally, we innovatively introduce a selective communication scheme in biology to decode mixed firing features using two descending neurons. We proceed to integrate such a neural circuit with a robot for avoidance control and achieve lower latency than conventional platforms. These results provide a foundation for implementing real brain-like systems driven by firing features with memristive neurons and put constructing high-order intelligent machines on the agenda.
Collapse
Affiliation(s)
- Yue Yang
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China
| | - Fangduo Zhu
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
| | - Xumeng Zhang
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China.
| | - Pei Chen
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
| | - Yongzhou Wang
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China
| | - Jiaxue Zhu
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China
| | - Yanting Ding
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
| | - Lingli Cheng
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China
| | - Chao Li
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China
| | - Hao Jiang
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
| | - Zhongrui Wang
- Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, 999077, China
| | - Peng Lin
- College of Computer Science and Technology, Zhejiang University, Zhejiang, 310027, China
| | - Tuo Shi
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China
| | - Ming Wang
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
| | - Qi Liu
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China.
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China.
| | - Ningsheng Xu
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
| | - Ming Liu
- State Key Laboratory of Integrated Chips and Systems, Frontier Institute of Chip and System, Fudan University, Shanghai, 200433, China
- Key Laboratory of Microelectronics Device & Integrated Technology, Institute of Microelectronics of Chinese Academy of Sciences, Beijing, 100029, China
| |
Collapse
|
2
|
Henkes A, Eshraghian JK, Wessels H. Spiking neural networks for nonlinear regression. ROYAL SOCIETY OPEN SCIENCE 2024; 11:231606. [PMID: 38699557 PMCID: PMC11062414 DOI: 10.1098/rsos.231606] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 10/25/2023] [Revised: 01/25/2024] [Accepted: 02/12/2024] [Indexed: 05/05/2024]
Abstract
Spiking neural networks (SNN), also often referred to as the third generation of neural networks, carry the potential for a massive reduction in memory and energy consumption over traditional, second-generation neural networks. Inspired by the undisputed efficiency of the human brain, they introduce temporal and neuronal sparsity, which can be exploited by next-generation neuromorphic hardware. Energy efficiency plays a crucial role in many engineering applications, for instance, in structural health monitoring. Machine learning in engineering contexts, especially in data-driven mechanics, focuses on regression. While regression with SNN has already been discussed in a variety of publications, in this contribution, we provide a novel formulation for its accuracy and energy efficiency. In particular, a network topology for decoding binary spike trains to real numbers is introduced, using the membrane potential of spiking neurons. Several different spiking neural architectures, ranging from simple spiking feed-forward to complex spiking long short-term memory neural networks, are derived. Since the proposed architectures do not contain any dense layers, they exploit the full potential of SNN in terms of energy efficiency. At the same time, the accuracy of the proposed SNN architectures is demonstrated by numerical examples, namely different material models. Linear and nonlinear, as well as history-dependent material models, are examined. While this contribution focuses on mechanical examples, the interested reader may regress any custom function by adapting the published source code.
Collapse
Affiliation(s)
- Alexander Henkes
- Computational Mechanics Group, ETH Zurich, Zurich, Switzerland
- Division Data-Driven Modeling of Mechanical Systems, Technical University Braunschweig, Braunschweig, Germany
| | - Jason K. Eshraghian
- Department of Electrical and Computer Engineering, University of California, Santa Cruz, CA, USA
| | - Henning Wessels
- Division Data-Driven Modeling of Mechanical Systems, Technical University Braunschweig, Braunschweig, Germany
| |
Collapse
|
3
|
D'Agostino S, Moro F, Torchet T, Demirağ Y, Grenouillet L, Castellani N, Indiveri G, Vianello E, Payvand M. DenRAM: neuromorphic dendritic architecture with RRAM for efficient temporal processing with delays. Nat Commun 2024; 15:3446. [PMID: 38658524 PMCID: PMC11043378 DOI: 10.1038/s41467-024-47764-w] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/14/2023] [Accepted: 04/11/2024] [Indexed: 04/26/2024] Open
Abstract
An increasing number of studies are highlighting the importance of spatial dendritic branching in pyramidal neurons in the neocortex for supporting non-linear computation through localized synaptic integration. In particular, dendritic branches play a key role in temporal signal processing and feature detection. This is accomplished thanks to coincidence detection (CD) mechanisms enabled by the presence of synaptic delays that align temporally disparate inputs for effective integration. Computational studies on spiking neural networks further highlight the significance of delays for achieving spatio-temporal pattern recognition with pure feed-forward neural networks, without the need of resorting to recurrent architectures. In this work, we present "DenRAM", the first realization of a feed-forward spiking neural network with dendritic compartments, implemented using analog electronic circuits integrated into a 130 nm technology node and coupled with Resistive Random Access Memory (RRAM) technology. DenRAM's dendritic circuits use RRAM devices to implement both delays and synaptic weights in the network. By configuring the RRAM devices to reproduce bio-realistic timescales, and by exploiting their heterogeneity we experimentally demonstrate DenRAM's ability to replicate synaptic delay profiles, and to efficiently implement CD for spatio-temporal pattern recognition. To validate the architecture, we conduct comprehensive system-level simulations on two representative temporal benchmarks, demonstrating DenRAM's resilience to analog hardware noise, and its superior accuracy compared to recurrent architectures with an equivalent number of parameters. DenRAM not only brings rich temporal processing capabilities to neuromorphic architectures, but also reduces the memory footprint of edge devices, warrants high accuracy on temporal benchmarks, and represents a significant step-forward in low-power real-time signal processing technologies.
Collapse
Affiliation(s)
- Simone D'Agostino
- Institute of Neuroinformatics, University of Zurich and ETH Zurich, Zurich, Switzerland
- CEA-Leti, Université Grenoble Alpes, Grenoble, France
| | - Filippo Moro
- Institute of Neuroinformatics, University of Zurich and ETH Zurich, Zurich, Switzerland
- CEA-Leti, Université Grenoble Alpes, Grenoble, France
| | - Tristan Torchet
- Institute of Neuroinformatics, University of Zurich and ETH Zurich, Zurich, Switzerland
| | - Yiğit Demirağ
- Institute of Neuroinformatics, University of Zurich and ETH Zurich, Zurich, Switzerland
| | | | | | - Giacomo Indiveri
- Institute of Neuroinformatics, University of Zurich and ETH Zurich, Zurich, Switzerland
| | | | - Melika Payvand
- Institute of Neuroinformatics, University of Zurich and ETH Zurich, Zurich, Switzerland.
| |
Collapse
|
4
|
Park M, Yang JY, Yeom MJ, Bae B, Baek Y, Yoo G, Lee K. An artificial neuromuscular junction for enhanced reflexes and oculomotor dynamics based on a ferroelectric CuInP 2S 6/GaN HEMT. SCIENCE ADVANCES 2023; 9:eadh9889. [PMID: 37738348 PMCID: PMC10516496 DOI: 10.1126/sciadv.adh9889] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 03/29/2023] [Accepted: 08/22/2023] [Indexed: 09/24/2023]
Abstract
A neuromuscular junction (NMJ) is a particularized synapse that activates muscle fibers for macro-motions, requiring more energy than computation. Emulating the NMJ is thus challenging owing to the need for both synaptic plasticity and high driving power to trigger motions. Here, we present an artificial NMJ using CuInP2S6 (CIPS) as a gate dielectric integrated with an AlGaN/GaN-based high-electron mobility transistor (HEMT). The ferroelectricity of the CIPS is coupled with the two-dimensional electron gas channel in the HEMT, providing a wide programmable current range of 6 picoampere per millimeter to 5 milliampere per millimeter. The large output current window of the CIPS/GaN ferroelectric HEMT (FeHEMT) allows for amplifier-less actuation, emulating the biological NMJ functions of actuation and synaptic plasticity. We also demonstrate the emulation of biological oculomotor dynamics, including in situ object tracking and enhanced stimulus responses, using the fabricated artificial NMJ. We believe that the CIPS/GaN FeHEMT offers a promising pathway for bioinspired robotics and neuromorphic vision.
Collapse
Affiliation(s)
- Minseong Park
- Department of Electrical and Computer Engineering, University of Virginia, Charlottesville, VA 22904, USA
| | - Jeong Yong Yang
- School of Electronic Engineering, Soongsil University, Seoul 06938, South Korea
| | - Min Jae Yeom
- School of Electronic Engineering, Soongsil University, Seoul 06938, South Korea
| | - Byungjoon Bae
- Department of Electrical and Computer Engineering, University of Virginia, Charlottesville, VA 22904, USA
| | - Yongmin Baek
- Department of Electrical and Computer Engineering, University of Virginia, Charlottesville, VA 22904, USA
| | - Geonwook Yoo
- School of Electronic Engineering, Soongsil University, Seoul 06938, South Korea
- Department of Intelligent Semiconductors, Soongsil University, Seoul 06978, South Korea
| | - Kyusang Lee
- Department of Electrical and Computer Engineering, University of Virginia, Charlottesville, VA 22904, USA
- Department of Material Science and Engineering, University of Virginia, Charlottesville, VA 22904, USA
| |
Collapse
|
5
|
Roy K, Lee JEY, Lee C. Thin-film PMUTs: a review of over 40 years of research. MICROSYSTEMS & NANOENGINEERING 2023; 9:95. [PMID: 37484500 PMCID: PMC10359338 DOI: 10.1038/s41378-023-00555-7] [Citation(s) in RCA: 10] [Impact Index Per Article: 10.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 01/14/2023] [Revised: 04/19/2023] [Accepted: 05/10/2023] [Indexed: 07/25/2023]
Abstract
Thin-film PMUTs have been important research topics among microultrasound experts, and a concise review on their research progress is reported herein. Through rigorous surveying, scrutinization, and perception, it has been determined that the work in this field began nearly 44 years ago with the primitive development of functional piezoelectric thin-film materials. To date, there are three major companies commercializing thin-film PMUTs on a bulk scale. This commercialization illustrates the extensive contributions made by more than 70 different centers, research institutes, and agencies across 4 different continents regarding the vast development of these devices' design, manufacturing, and function. This review covers these important contributions in a short yet comprehensive manner; in particular, this paper educates readers about the global PMUT outlook, their governing design principles, their manufacturing methods, nonconventional yet useful PMUT designs, and category-wise applications. Crucial comparison charts of thin-film piezoelectric material used in PMUTs, and their categorically targeted applications are depicted and discussed to enlighten any MEMS designer who plans to work with PMUTs. Moreover, each relevant section features clear future predictions based on the author's past knowledge and expertise in this field of research and on the findings of a careful literature survey. In short, this review is a one-stop time-efficient guide for anyone interested in learning about these small devices.
Collapse
Affiliation(s)
- Kaustav Roy
- Department of Electrical and Computer Engineering, National University of Singapore, Singapore, 117583 Singapore
- Center for Intelligent Sensor and MEMS (CISM), National University of Singapore, Singapore, 117608 Singapore
| | | | - Chengkuo Lee
- Department of Electrical and Computer Engineering, National University of Singapore, Singapore, 117583 Singapore
- Center for Intelligent Sensor and MEMS (CISM), National University of Singapore, Singapore, 117608 Singapore
| |
Collapse
|
6
|
Crepaldi M, Mohan C, Garofalo E, Adamatzky A, Szaciłowski K, Chiolerio A. Experimental Demonstration of In-Memory Computing in a Ferrofluid System. ADVANCED MATERIALS (DEERFIELD BEACH, FLA.) 2023; 35:e2211406. [PMID: 36919899 DOI: 10.1002/adma.202211406] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 12/06/2022] [Revised: 02/27/2023] [Indexed: 06/09/2023]
Abstract
Magnetic fluids are excellent candidates for several important research fields including energy harvesting, biomedical applications, soft robotics, and exploration. However, notwithstanding relevant advancements such as shape reconfigurability, that have been demonstrated, there is no evidence for their computing capability, including the emulation of synaptic functions, which requires complex non-linear dynamics. Here, it is experimentally demonstrated that a Fe3 O4 water-based ferrofluid (FF) can perform electrical analogue computing and be programmed using quasi direct current (DC) signals and read at radio frequency (RF) mode. Features have been observed in all respects attributable to a memristive behavior, featuring both short and long-term information storage capacity and plasticity. The colloid is capable of classifying digits of a 8 × 8 pixel dataset using a custom in-memory signal processing scheme, and through physical reservoir computing by training a readout layer. These findings demonstrate the feasibility of in-memory computing using an amorphous FF system in a liquid aggregation state. This work poses the basis for the exploitation of a FF colloid as both an in-memory computing device and as a full-electric liquid computer thanks to its fluidity and the reported complex dynamics, via probing read-out and programming ports.
Collapse
Affiliation(s)
- Marco Crepaldi
- Electronic Design Laboratory, Istituto Italiano di Tecnologia, Via Melen 83, Genova, Liguria, 16152, Italy
| | - Charanraj Mohan
- Electronic Design Laboratory, Istituto Italiano di Tecnologia, Via Melen 83, Genova, Liguria, 16152, Italy
| | - Erik Garofalo
- Bioinspired Soft Robotics, Istituto Italiano di Tecnologia, Via Morego 30, Genova, Liguria, 16163, Italy
| | - Andrew Adamatzky
- Unconventional Computing Laboratory, University of West England, Frenchay Campus, Coldharbour Ln, Bristol, Bristol, BS16 1QY, UK
| | - Konrad Szaciłowski
- Academic Centre for Materials and Nanotechnology, AGH University of Technology, 30 Mickiewicza Avenue, Kraków, 30-059, Poland
| | - Alessandro Chiolerio
- Bioinspired Soft Robotics, Istituto Italiano di Tecnologia, Via Morego 30, Genova, Liguria, 16163, Italy
- Unconventional Computing Laboratory, University of West England, Frenchay Campus, Coldharbour Ln, Bristol, Bristol, BS16 1QY, UK
| |
Collapse
|
7
|
Schmid D, Oess T, Neumann H. Listen to the Brain-Auditory Sound Source Localization in Neuromorphic Computing Architectures. SENSORS (BASEL, SWITZERLAND) 2023; 23:s23094451. [PMID: 37177655 PMCID: PMC10181665 DOI: 10.3390/s23094451] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 03/06/2023] [Revised: 04/23/2023] [Accepted: 04/27/2023] [Indexed: 05/15/2023]
Abstract
Conventional processing of sensory input often relies on uniform sampling leading to redundant information and unnecessary resource consumption throughout the entire processing pipeline. Neuromorphic computing challenges these conventions by mimicking biology and employing distributed event-based hardware. Based on the task of lateral auditory sound source localization (SSL), we propose a generic approach to map biologically inspired neural networks to neuromorphic hardware. First, we model the neural mechanisms of SSL based on the interaural level difference (ILD). Afterward, we identify generic computational motifs within the model and transform them into spike-based components. A hardware-specific step then implements them on neuromorphic hardware. We exemplify our approach by mapping the neural SSL model onto two platforms, namely the IBM TrueNorth Neurosynaptic System and SpiNNaker. Both implementations have been tested on synthetic and real-world data in terms of neural tunings and readout characteristics. For synthetic stimuli, both implementations provide a perfect readout (100% accuracy). Preliminary real-world experiments yield accuracies of 78% (TrueNorth) and 13% (SpiNNaker), RMSEs of 41∘ and 39∘, and MAEs of 18∘ and 29∘, respectively. Overall, the proposed mapping approach allows for the successful implementation of the same SSL model on two different neuromorphic architectures paving the way toward more hardware-independent neural SSL.
Collapse
Affiliation(s)
- Daniel Schmid
- Institute of Neural Information Processing, Ulm University, James-Franck-Ring, 89081 Ulm, Germany
| | - Timo Oess
- Bernstein Center Freiburg, University of Freiburg, Hansastr. 9a, 79104 Freiburg im Breisgau, Germany
| | - Heiko Neumann
- Institute of Neural Information Processing, Ulm University, James-Franck-Ring, 89081 Ulm, Germany
| |
Collapse
|
8
|
Wang C, Shi G, Qiao F, Lin R, Wu S, Hu Z. Research progress in architecture and application of RRAM with computing-in-memory. NANOSCALE ADVANCES 2023; 5:1559-1573. [PMID: 36926563 PMCID: PMC10012847 DOI: 10.1039/d3na00025g] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 01/11/2023] [Accepted: 02/04/2023] [Indexed: 06/18/2023]
Abstract
The development of new technologies has led to an explosion of data, while the computation ability of traditional computers is approaching its upper limit. The dominant system architecture is the von Neumann architecture, with the processing and storage units working independently. The data migrate between them via buses, reducing computing speed and increasing energy loss. Research is underway to increase computing power, such as developing new chips and adopting new system architectures. Computing-in-memory (CIM) technology allows data to be computed directly on the memory, changing the current computation-centric architecture and designing a new storage-centric architecture. Resistive random access memory (RRAM) is one of the advanced memories which has appeared in recent years. RRAM can change its resistance with electrical signals at both ends, and the state will be preserved after power-down. It has potential in logic computing, neural networks, brain-like computing, and fused technology of sense-storage-computing. These advanced technologies promise to break the performance bottleneck of traditional architectures and dramatically increase computing power. This paper introduces the basic concepts of computing-in-memory technology and the principle and applications of RRAM and finally gives a conclusion about these new technologies.
Collapse
Affiliation(s)
- Chenyu Wang
- College of Mechanical and Electrical Engineering, China Jiliang University Hangzhou China
| | - Ge Shi
- College of Mechanical and Electrical Engineering, China Jiliang University Hangzhou China
| | - Fei Qiao
- Dept of Electronic Engineering, Tsinghua University Beijing 310018 People's Republic of China
| | - Rubin Lin
- College of Mechanical and Electrical Engineering, China Jiliang University Hangzhou China
| | - Shien Wu
- College of Mechanical and Electrical Engineering, China Jiliang University Hangzhou China
| | - Zenan Hu
- College of Mechanical and Electrical Engineering, China Jiliang University Hangzhou China
| |
Collapse
|