1
|
Youn S, Hwang Y, Kim TH, Kim S, Hwang H, Park J, Kim H. Threshold learning algorithm for memristive neural network with binary switching behavior. Neural Netw 2024; 176:106355. [PMID: 38759411 DOI: 10.1016/j.neunet.2024.106355] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/09/2023] [Revised: 02/04/2024] [Accepted: 04/29/2024] [Indexed: 05/19/2024]
Abstract
On-chip learning is an effective method for adjusting artificial neural networks in neuromorphic computing systems by considering hardware intrinsic properties. However, it faces challenges due to hardware nonidealities, such as the nonlinearity of potentiation and depression and limitations on fine weight adjustment. In this study, we propose a threshold learning algorithm for a variation-tolerant ternary neural network in a memristor crossbar array. This algorithm utilizes two tightly separated resistance states in memristive devices to represent weight values. The high-resistance state (HRS) and low-resistance state (LRS) defined as read current of < 0.1 μA and > 1 μA, respectively, were successfully programmed in a 32 × 32 crossbar array, and exhibited half-normal distributions due to the programming method. To validate our approach experimentally, a 64 × 10 single-layer fully connected network were trained in the fabricated crossbar for an 8 × 8 MNIST dataset using the threshold learning algorithm, where the weight value is updated when a gradient determined by backpropagation exceeds a threshold value. Thanks to the large margin between the two states of the memristor, we observed only a 0.42 % drop in classification accuracy compared to the baseline network results. The threshold learning algorithm is expected to alleviate the programming burden and be utilized in variation-tolerant neuromorphic architectures.
Collapse
Affiliation(s)
- Sangwook Youn
- Division of Materials Science and Engineering, Seoul 04763, Korea
| | - Yeongjin Hwang
- Department of Electrical and Computer Engineering, Inha University, Incheon 22212, Korea
| | - Tae-Hyeon Kim
- Department of Semiconductor Engineering, Seoul National University of Science and Technology, Seoul 01811, Korea
| | - Sungjoon Kim
- Department of AI Semiconductor Engineering, Korea University, Sejong 30019, Korea
| | - Hwiho Hwang
- Division of Materials Science and Engineering, Seoul 04763, Korea
| | - Jinwoo Park
- Division of Materials Science and Engineering, Seoul 04763, Korea
| | - Hyungjin Kim
- Division of Materials Science and Engineering, Seoul 04763, Korea.
| |
Collapse
|
2
|
Galloni AR, Yuan Y, Zhu M, Yu H, Bisht RS, Wu CTM, Grienberger C, Ramanathan S, Milstein AD. Neuromorphic one-shot learning utilizing a phase-transition material. Proc Natl Acad Sci U S A 2024; 121:e2318362121. [PMID: 38630718 PMCID: PMC11047090 DOI: 10.1073/pnas.2318362121] [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: 10/20/2023] [Accepted: 03/25/2024] [Indexed: 04/19/2024] Open
Abstract
Design of hardware based on biological principles of neuronal computation and plasticity in the brain is a leading approach to realizing energy- and sample-efficient AI and learning machines. An important factor in selection of the hardware building blocks is the identification of candidate materials with physical properties suitable to emulate the large dynamic ranges and varied timescales of neuronal signaling. Previous work has shown that the all-or-none spiking behavior of neurons can be mimicked by threshold switches utilizing material phase transitions. Here, we demonstrate that devices based on a prototypical metal-insulator-transition material, vanadium dioxide (VO2), can be dynamically controlled to access a continuum of intermediate resistance states. Furthermore, the timescale of their intrinsic relaxation can be configured to match a range of biologically relevant timescales from milliseconds to seconds. We exploit these device properties to emulate three aspects of neuronal analog computation: fast (~1 ms) spiking in a neuronal soma compartment, slow (~100 ms) spiking in a dendritic compartment, and ultraslow (~1 s) biochemical signaling involved in temporal credit assignment for a recently discovered biological mechanism of one-shot learning. Simulations show that an artificial neural network using properties of VO2 devices to control an agent navigating a spatial environment can learn an efficient path to a reward in up to fourfold fewer trials than standard methods. The phase relaxations described in our study may be engineered in a variety of materials and can be controlled by thermal, electrical, or optical stimuli, suggesting further opportunities to emulate biological learning in neuromorphic hardware.
Collapse
Affiliation(s)
- Alessandro R. Galloni
- Department of Neuroscience and Cell Biology, Robert Wood Johnson Medical School, Rutgers, The State University of New Jersey, Piscataway, NJ08854
- Center for Advanced Biotechnology and Medicine, Rutgers, The State University of New Jersey, Piscataway, NJ08854
| | - Yifan Yuan
- Department of Electrical and Computer Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ08854
| | - Minning Zhu
- Department of Electrical and Computer Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ08854
| | - Haoming Yu
- School of Materials Engineering, Purdue University, West Lafayette, IN47907
| | - Ravindra S. Bisht
- Department of Electrical and Computer Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ08854
| | - Chung-Tse Michael Wu
- Department of Electrical and Computer Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ08854
| | - Christine Grienberger
- Department of Neuroscience, Brandeis University, Waltham, MA02453
- Department of Biology and Volen National Center for Complex Systems, Brandeis University, Waltham, MA02453
| | - Shriram Ramanathan
- Department of Electrical and Computer Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ08854
| | - Aaron D. Milstein
- Department of Neuroscience and Cell Biology, Robert Wood Johnson Medical School, Rutgers, The State University of New Jersey, Piscataway, NJ08854
- Center for Advanced Biotechnology and Medicine, Rutgers, The State University of New Jersey, Piscataway, NJ08854
| |
Collapse
|
3
|
Kim K, Song MS, Hwang H, Hwang S, Kim H. A comprehensive review of advanced trends: from artificial synapses to neuromorphic systems with consideration of non-ideal effects. Front Neurosci 2024; 18:1279708. [PMID: 38660225 PMCID: PMC11042536 DOI: 10.3389/fnins.2024.1279708] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/18/2023] [Accepted: 03/14/2024] [Indexed: 04/26/2024] Open
Abstract
A neuromorphic system is composed of hardware-based artificial neurons and synaptic devices, designed to improve the efficiency of neural computations inspired by energy-efficient and parallel operations of the biological nervous system. A synaptic device-based array can compute vector-matrix multiplication (VMM) with given input voltage signals, as a non-volatile memory device stores the weight information of the neural network in the form of conductance or capacitance. However, unlike software-based neural networks, the neuromorphic system unavoidably exhibits non-ideal characteristics that can have an adverse impact on overall system performance. In this study, the characteristics required for synaptic devices and their importance are discussed, depending on the targeted application. We categorize synaptic devices into two types: conductance-based and capacitance-based, and thoroughly explore the operations and characteristics of each device. The array structure according to the device structure and the VMM operation mechanism of each structure are analyzed, including recent advances in array-level implementation of synaptic devices. Furthermore, we reviewed studies to minimize the effect of hardware non-idealities, which degrades the performance of hardware neural networks. These studies introduce techniques in hardware and signal engineering, as well as software-hardware co-optimization, to address these non-idealities through compensation approaches.
Collapse
Affiliation(s)
- Kyuree Kim
- Department of Electrical and Computer Engineering, Inha University, Incheon, Republic of Korea
| | - Min Suk Song
- Division of Nanoscale Semiconductor Engineering, Hanyang University, Seoul, Republic of Korea
| | - Hwiho Hwang
- Division of Materials Science and Engineering, Hanyang University, Seoul, Republic of Korea
| | - Sungmin Hwang
- Department of AI Semiconductor Engineering, Korea University, Sejong, Republic of Korea
| | - Hyungjin Kim
- Division of Materials Science and Engineering, Hanyang University, Seoul, Republic of Korea
| |
Collapse
|
4
|
Youn S, Lee J, Kim S, Park J, Kim K, Kim H. Programmable Threshold Logic Implementations in a Memristor Crossbar Array. NANO LETTERS 2024; 24:3581-3589. [PMID: 38471119 DOI: 10.1021/acs.nanolett.3c04073] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 03/14/2024]
Abstract
In this study, we demonstrate the implementation of programmable threshold logics using a 32 × 32 memristor crossbar array. Thanks to forming-free characteristics obtained by the annealing process, its accurate programming characteristics are presented by a 256-level grayscale image. By simultaneous subtraction between weighted sum and threshold values with a differential pair in an opposite way, 3-input and 4-input Boolean logics are implemented in the crossbar without additional reference bias. Also, we verify a full-adder circuit and analyze its fidelity, depending on the device programming accuracy. Lastly, we successfully implement a 4-bit ripple carry adder in the crossbar and achieve reliable operations by read-based logic operations. Compared to stateful logic driven by device switching, a 4-bit ripple carry adder on a memristor crossbar array can perform more reliably in fewer steps thanks to its read-based parallel logic operation.
Collapse
Affiliation(s)
- Sangwook Youn
- Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Korea
| | - Jungjin Lee
- Department of Electrical and Computer Engineering, Inha University, Incheon 22212, Korea
| | - Sungjoon Kim
- Department of Electrical and Computer Engineering, Seoul National University, Seoul 08826, Korea
| | - Jinwoo Park
- Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Korea
| | - Kyuree Kim
- Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Korea
| | - Hyungjin Kim
- Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Korea
- Department of Electronic Engineering, Hanyang University, Seoul 04763, Korea
| |
Collapse
|
5
|
Kwon JY, Kim JE, Kim JS, Chun SY, Soh K, Yoon JH. Artificial sensory system based on memristive devices. EXPLORATION (BEIJING, CHINA) 2024; 4:20220162. [PMID: 38854486 PMCID: PMC10867403 DOI: 10.1002/exp.20220162] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 07/20/2023] [Accepted: 10/16/2023] [Indexed: 06/11/2024]
Abstract
In the biological nervous system, the integration and cooperation of parallel system of receptors, neurons, and synapses allow efficient detection and processing of intricate and disordered external information. Such systems acquire and process environmental data in real-time, efficiently handling complex tasks with minimal energy consumption. Memristors can mimic typical biological receptors, neurons, and synapses by implementing key features of neuronal signal-processing functions such as selective adaption in receptors, leaky integrate-and-fire in neurons, and synaptic plasticity in synapses. External stimuli are sensitively detected and filtered by "artificial receptors," encoded into spike signals via "artificial neurons," and integrated and stored through "artificial synapses." The high operational speed, low power consumption, and superior scalability of memristive devices make their integration with high-performance sensors a promising approach for creating integrated artificial sensory systems. These integrated systems can extract useful data from a large volume of raw data, facilitating real-time detection and processing of environmental information. This review explores the recent advances in memristor-based artificial sensory systems. The authors begin with the requirements of artificial sensory elements and then present an in-depth review of such elements demonstrated by memristive devices. Finally, the major challenges and opportunities in the development of memristor-based artificial sensory systems are discussed.
Collapse
Affiliation(s)
- Ju Young Kwon
- Electronic Materials Research CenterKorea Institute of Science and Technology (KIST)SeoulRepublic of Korea
| | - Ji Eun Kim
- Electronic Materials Research CenterKorea Institute of Science and Technology (KIST)SeoulRepublic of Korea
- Department of Materials Science and EngineeringKorea UniversitySeoulRepublic of Korea
| | - Jong Sung Kim
- Electronic Materials Research CenterKorea Institute of Science and Technology (KIST)SeoulRepublic of Korea
- Department of Materials Science and EngineeringKorea UniversitySeoulRepublic of Korea
| | - Suk Yeop Chun
- Electronic Materials Research CenterKorea Institute of Science and Technology (KIST)SeoulRepublic of Korea
- KU‐KIST Graduate School of Converging Science and TechnologyKorea UniversitySeoulRepublic of Korea
| | - Keunho Soh
- Electronic Materials Research CenterKorea Institute of Science and Technology (KIST)SeoulRepublic of Korea
- Department of Materials Science and EngineeringKorea UniversitySeoulRepublic of Korea
| | - Jung Ho Yoon
- Electronic Materials Research CenterKorea Institute of Science and Technology (KIST)SeoulRepublic of Korea
| |
Collapse
|
6
|
Park J, Kim S, Song MS, Youn S, Kim K, Kim TH, Kim H. Implementation of Convolutional Neural Networks in Memristor Crossbar Arrays with Binary Activation and Weight Quantization. ACS APPLIED MATERIALS & INTERFACES 2024; 16:1054-1065. [PMID: 38163259 DOI: 10.1021/acsami.3c13775] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 01/03/2024]
Abstract
We propose a hardware-friendly architecture of a convolutional neural network using a 32 × 32 memristor crossbar array having an overshoot suppression layer. The gradual switching characteristics in both set and reset operations enable the implementation of a 3-bit multilevel operation in a whole array that can be utilized as 16 kernels. Moreover, a binary activation function mapped to the read voltage and ground is introduced to evaluate the result of training with a boundary of 0.5 and its estimated gradient. Additionally, we adopt a fixed kernel method, where inputs are sequentially applied to a crossbar array with a differential memristor pair scheme, reducing unused cell waste. The binary activation has robust characteristics against device state variations, and a neuron circuit is experimentally demonstrated on a customized breadboard. Thanks to the analogue switching characteristics of the memristor device, the accurate vector-matrix multiplication (VMM) operations can be experimentally demonstrated by combining sequential inputs and the weights obtained through tuning operations in the crossbar array. In addition, the feature images extracted by VMM during the hardware inference operations on 100 test samples are classified, and the classification performance by off-chip training is compared with the software results. Finally, inference results depending on the tolerance are statistically verified through several tuning cycles.
Collapse
Affiliation(s)
- Jinwoo Park
- Department of Electrical and Computer Engineering, Inha University, Incheon 22212, Korea
| | - Sungjoon Kim
- Department of Electrical and Computer Engineering, Seoul National University, Seoul 08826, Korea
| | - Min Suk Song
- Department of Electrical and Computer Engineering, Inha University, Incheon 22212, Korea
| | - Sangwook Youn
- Department of Electrical and Computer Engineering, Inha University, Incheon 22212, Korea
| | - Kyuree Kim
- Department of Electrical and Computer Engineering, Inha University, Incheon 22212, Korea
| | - Tae-Hyeon Kim
- Department of Semiconductor Engineering, Seoul National University of Science and Technology, Seoul 01811, Korea
| | - Hyungjin Kim
- Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Korea
| |
Collapse
|