检索结果-内蒙古大学图书馆

Random resistive memory-based deep extreme point learning machine for unified visual processing

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Shaocong Gao, Yizhao Li, Yi Zhang, Woyu Yu, Yifei Wang, Bo Lin, Ning Chen, Hegan Zhang, Yue Jiang, Yang Wang, Dingchen Chen, Jia Dai, Peng Jiang, Hao Lin, Peng Zhang, Xumeng Qi, Xiaojuan Xu, Xiaoxin So, Hayden Wang, Zhongrui Shang, Dashan Liu, Qi Cheng, Kwang-Ting Liu, Ming Department of Electrical and Electronic Engineering The University of Hong Kong Hong Kong Key Laboratory of Microelectronic Devices & Integrated Technology Institute of Microelectronics Chinese Academy of Sciences Beijing100029 China ACCESS - AI Chip Center for Emerging Smart Systems InnoHK Centers Hong Kong Science Park Hong Kong University of Chinese Academy of Sciences Beijing100049 China Frontier Institute of Chip and System Fudan University Shanghai200433 China College of Computer Science and Technology Zhejiang University Zhejiang310027 China Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology Hong Kong

Visual sensors, including 3D LiDAR, neuromorphic DVS sensors, and conventional frame cameras, are increasingly integrated into edge-side intelligent machines. Realizing intensive multi-sensory data analysis directly on edge intelligent machines is crucial for numerous emerging edge applications, such as augmented and virtual reality and unmanned aerial vehicles, which necessitates unified data representation, unprecedented hardware energy efficiency and rapid model training. However, multi-sensory data are intrinsically heterogeneous, causing significant complexity in the system development for edge-side intelligent machines. In addition, the performance of conventional digital hardware is limited by the physically separated processing and memory units, known as the von Neumann bottleneck, and the physical limit of transistor scaling, which contributes to the slowdown of Moore's law. These limitations are further intensified by the tedious training of models with ever-increasing sizes. In this study, we propose a novel hardware-software co-design, random resistive memory-based deep extreme point learning machine (DEPLM), that offers efficient unified point set analysis. Data-wise, the multi-sensory data are unified as point sets and can be processed universally. Software-wise, most weights of deep extreme point learning machines are exempted from training, which significantly reduce training complexity. Hardware-wise, nanoscale resistive memory not only enables collocation of memory and processing, mitigating the von Neumann bottleneck and the slowdown of Moore's law, but also leverages the inherent programming stochasticity for generating the random and sparse weights of the DEPLM, which also lessens the impact of read noise. We demonstrate the system's versatility across various data modalities and two different learning tasks. Compared to a conventional digital hardware-based system, our co-design system achieves 5.90×, 21.04×, and 15.79× energy efficiency improv

关键词： Cost reduction

Causally-informed Deep Learning to Improve Climate Models and Projections

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Iglesias-Suarez, Fernando Gentine, Pierre Solino-Fernandez, Breixo Beucler, Tom Pritchard, Michael Runge, Jakob Eyring, Veronika Institute of Atmospheric Physics Oberpfaffenhofen Germany Columbia University New York United States Science and Technology Center Columbia University New York United States University of Lausanne Institute of Earth Surface Dynamics Lausanne Switzerland University of California Department of Earth System Science Irvine United States NVIDIA Corporation Santa Clara United States Institute of Data Science Jena Germany Technische Universität Berlin Institute of Computer Engineering and Microelectronics Berlin Germany Bremen Germany

Climate models are essential to understand and project climate change, yet long-standing biases and uncertainties in their projections remain. This is largely associated with the representation of subgrid-scale processes, particularly clouds and convection. Deep learning can learn these subgrid-scale processes from computationally expensive storm-resolving models while retaining many features at a fraction of computational cost. Yet, climate simulations with embedded neural network parameterizations are still challenging and highly depend on the deep learning solution. This is likely associated with spurious nonphysical correlations learned by the neural networks due to the complexity of the physical dynamical system. Here, we show that the combination of causality with deep learning helps removing spurious correlations and optimizing the neural network algorithm. To resolve this, we apply a causal discovery method to unveil causal drivers in the set of input predictors of atmospheric subgrid-scale processes of a superparameterized climate model in which deep convection is explicitly resolved. The resulting causally-informed neural networks are coupled to the climate model, hence, replacing the superparameterization and radiation scheme. We show that the climate simulations with causally-informed neural network parameterizations retain many convection-related properties and accurately generate the climate of the original high-resolution climate model, while retaining similar generalization capabilities to unseen climates compared to the non-causal approach. The combination of causal discovery and deep learning is a new and promising approach that leads to stable and more trustworthy climate simulations and paves the way towards more physically-based causal deep learning approaches also in other scientific disciplines. © 2023, CC BY-NC-ND.

关键词： Climate models

Resistive memory-based zero-shot liquid state machine for multimodal event data learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lin, Ning Wang, Shaocong Li, Yi Wang, Bo Shi, Shuhui He, Yangu Zhang, Woyu Yu, Yifei Zhang, Yue Zhang, Xinyuan Wong, Kwunhang Wang, Songqi Chen, Xiaoming Jiang, Hao Zhang, Xumeng Lin, Peng Xu, Xiaoxin Qi, Xiaojuan Wang, Zhongrui Shang, Dashan Liu, Qi Liu, Ming Department of Electrical and Electronic Engineering the University of Hong Kong Hong Kong Key Lab of Fabrication Technologies for Integrated Circuits and Key Laoratory of Microelectronic Devices and Integrated Technology Institute of Microelectronics the Chinese Academy of Sciences Beijing100029 China The School of Microelectronics Southern University of Science and Technology Shenzhen518055 China ACCESS – AI Chip Center for Emerging Smart Systems InnoHK Centers Hong Kong Science Park Hong Kong University of Chinese Academy of Sciences Beijing100049 China Frontier Institute of Chip and System Fudan University Shanghai200433 China College of Computer Science and Technology Zhejiang University Hangzhou310027 China Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China

The human brain is a complex spiking neural network (SNN), capable of learning multimodal signals in a zero-shot manner by generalizing existing knowledge. Remarkably, it maintains minimal power consumption through event-based signal propagation. However, replicating the human brain in neuromorphic hardware presents both hardware and software challenges. Hardware limitations, such as the slowdown of Moore’s law and Von Neumann bottleneck, hinder the efficiency of digital computers. Additionally, SNNs are characterized by their software training complexities. To this end, we propose a hardware-software co-design on a 40 nm 256 Kb in-memory computing macro that physically integrates a fixed and random liquid state machine (LSM) SNN encoder with trainable artificial neural network (ANN) projections. We showcase the zero-shot LSM-based learning of multimodal events on the N-MNIST and N-TIDIGITS datasets, including visual and audio data association, as well as neural and visual data alignment for brain-machine interfaces. Our co-design achieves classification accuracy comparable to fully optimized software models, resulting in a 152.83 and 393.07-fold reduction in training costs compared to SOTA contrastive language-image pre-training (CLIP) and Prototypical networks, and a 23.34 and 160-fold improvement in energy efficiency compared to cutting-edge digital hardware, respectively. These proof-of-principle prototypes demonstrate zero-shot multimodal events learning capability for emerging efficient and compact neuromorphic hardware. © 2023, CC BY-NC-SA.

关键词： Image enhancement

Hybrid memristor-CMOS neurons for in-situ learning in fully hardware memristive spiking neural networks

学校读者我要写书评

暂无评论

Science Bulletin 2021年第16期66卷 1624-1633,M0003页

作者： Xumeng Zhang Jian Lu Zhongrui Wang Rui Wang Jinsong Wei Tuo Shi Chunmeng Dou Zuheng Wu Jiaxue Zhu Dashan Shang Guozhong Xing Mansun Chan Qi Liu Ming Liu Frontier Institute of Chip and System Fudan UniversityShanghai 200433China Key Laboratory of Microelectronic Devices&Integrated Technology Institute of MicroelectronicsChinese Academy of SciencesBeijing 100029China University of Chinese Academy of Sciences Beijing 100049China Department of Electrical and Electronic Engineering the University of Hong KongHong KongChina Zhejiang Laboratory Hangzhou 311122China Department of Electronic and Computer Engineering the Hong Kong University of Science and TechnologyHong KongChina

Spiking neural network,inspired by the human brain,consisting of spiking neurons and plastic synapses,is a promising solution for highly efficient data processing in neuromorphic ***,memristor-based neurons and synapses are becoming intriguing candidates to build spiking neural networks in hardware,owing to the close resemblance between their device dynamics and the biological ***,the functionalities of memristor-based neurons are currently very limited,and a hardware demonstration of fully memristor-based spiking neural networks supporting in-situ learning is very ***,a hybrid spiking neuron combining a memristor with simple digital circuits is designed and implemented in hardware to enhance neuron *** hybrid neuron with memristive dynamics not only realizes the basic leaky integrate-and-fire neuron function but also enables the in-situ tuning of the connected synaptic ***,a fully hardware spiking neural network with the hybrid neurons and memristive synapses is experimentally demonstrated for the first time,and in-situ Hebbian learning is achieved with this *** work opens up a way towards the implementation of spiking neurons,supporting in-situ learning for future neuromorphic computing systems.

关键词： Memristor Hybrid neuron In-situ learning Fully hardware Spiking neural network

Ultrafast non-volatile flash memory based on van der Waals heterostructures

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Liu, Lan Ding, Yi Li, Jiayi Liu, Chunsen Zhou, Peng State Key Laboratory of ASIC and System School of Microelectronics Fudan University Shanghai200433 China School of Computer Science Fudan University Shanghai200433 China

Flash memory has become a ubiquitous solid-state memory device, it is widely used in portable digital devices, computers, and enterprise applications. The development of the information age has put forward higher requirements for memory speed and retention performance. Here, we demonstrate an ultrafast non-volatile memory based on MoS2/h-BN/multi-layer graphene (MLG) van der Waals heterostructures, which has an atomic-level flat interface and achieves ultrafast writing/erasing speed (~20 ns), surpassing the reported state-of-the-art flash memory (~100 ns). The ultrafast flash memory could lay the foundation for the next-generation of high-speed non-volatile memory. Copyright © 2020, The Authors. All rights reserved.

关键词： Layered semiconductors

Dynamic neural network with memristive CIM and CAM for 2D and 3D vision

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhang, Yue Zhang, Woyu Wang, Shaocong Lin, Ning Yu, Yifei He, Yangu Wang, Bo Jiang, Hao Lin, Peng Xu, Xiaoxin Qi, Xiaojuan Wang, Zhongrui Zhang, Xumeng Shang, Dashan Liu, Qi Cheng, Kwang-Ting Liu, Ming Department of Electrical and Electronic Engineering The University of Hong Kong Hong Kong Key Laboratory of Microelectronic Devices & Integrated Technology Institute of Microelectronics Chinese Academy of Sciences Beijing100049 China ACCESS – AI Chip Center for Emerging Smart Systems InnoHK Centers Hong Kong Science Park Hong Kong Institute of the Mind The University of Hong Kong Hong Kong Key Lab of Fabrication Technologies for Integrated Circuits Chinese Academy of Sciences Beijing100049 China State Key Laboratory of Integrated Chips and Systems Frontier Institute of Chip and System Fudan University Shanghai200433 China University of Chinese Academy of Sciences Beijing100049 China College of Computer Science and Technology Zhejiang University Zhejiang310027 China Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology Hong Kong

The brain is dynamic, associative and efficient. It reconfigures by associating the inputs with past experiences, with fused memory and processing. In contrast, AI models are static, unable to associate inputs with past experiences, and run on digital computers with physically separated memory and processing. We propose a hardware-software co-design, a semantic memory-based dynamic neural network (DNN) using memristor. The network associates incoming data with the past experience stored as semantic vectors. The network and the semantic memory are physically implemented on noise-robust ternary memristor-based Computing-In-Memory (CIM) and Content-Addressable Memory (CAM) circuits, respectively. We validate our co-designs, using a 40nm memristor macro, on ResNet and PointNet++ for classifying images and 3D points from the MNIST and ModelNet datasets, which not only achieves accuracy on par with software but also a 48.1% and 15.9% reduction in computational budget. Moreover, it delivers a 77.6% and 93.3% reduction in energy consumption. © 2024, CC BY.

关键词： Memristors

Information Freshness in A Dual Monitoring system

学校读者我要写书评

暂无评论

Information Freshness in A Dual Monitoring System

GLOBECOM 2022 - 2022 IEEE Global Communications Conference

作者： Dapeng Deng Zhengchuan Chen Howard H. Yang Nikolaos Pappas Limei Hu Min Wang Yunjian Jia Tony Q.S. Quek School of Microelectronics and Communication Engineering Chongqing University Chongqing China National Mobile Communications Research Laboratory Southeast University Nanjing China Zhejiang University/University of Illinois at Urbana-Champaign Institute Zhejiang University Haining China Department of Computer and Information Science Linkoping University Linkoping Sweden School of Optoelectronics Engineering Chongqing University of Posts and Telecommunications Chongqing China Information System Technology and Design Pillar Singapore University of Technology and Design Singapore

ISBN: (纸本)9781665435413

We study the average age of information (AoI) and peak AoI (PAoI) of a dual-queue status update system that monitors a common stochastic process. We capture the state transition characteristics of the considered system by establishing a Markov chain. Using the state flow graph analysis method, we derive closed-form expressions of the average peak age of information (PAoI) and the average age of information (AoI) for the dual-queue update system. The numerical results show that compared with the single-queue update system, the average PAoI of the dual-queue update system is reduced by 33.5% and the average AoI dropped by 37.5%.

关键词： Closed-form solutions Markov processes Reliability theory Information age Sensor systems Real-time systems Sensors

Graphene-based all-optical modulators

学校读者我要写书评

暂无评论

Frontiers of Optoelectronics 2020年第2期13卷 114-128,F0004页

作者： Chuyu ZHONG Junying LI Hongtao LIN Key Laboratory of Micro-Nano Electronics and Smart System of Zhejiang Province College of Information Science and Electronic EngineeringZhejiang UniversityHangzhou 310027China School of Microelectronics Zhejiang UniversityHangzhou 310027China College of Optical-Electrical and Computer Engineering University of Shanghai for Science and TechnologyShanghai 200093China

All-optical devices,which are utilized to process optical signals without electro-optical conversion,play an essential role in the next generation ultrafast,ultralow power-consumption optical information processing *** satisfy the performance requirement,nonlinear optical materials that are associated with fast response,high nonlinearity,broad wavelength operation,low optical loss,low fabrication cost,and integration compatibility with optical components are *** is a promising candidate,particularly considering its electrically or optically tunable optical properties,ultrafast large nonlinearity,and high integration compatibility with various *** far,three all-optical modulation systems utilize graphene,namely free-space modulators,fiber-based modulators,and on-chip *** paper aims to provide a broad view of state-of-the-art researches on the graphene-based all-optical modulation *** performances of different devices are reviewed and compared to present a comprehensive analysis and perspective of graphene-based all-optical modulation devices.

关键词： graphene saturable absorption low power consumption all-optical modulation

Resistive Memory-based Neural Differential Equation Solver for Score-based Diffusion Model

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yang, Jichang Chen, Hegan Chen, Jia Wang, Songqi Wang, Shaocong Yu, Yifei Chen, Xi Wang, Bo Zhang, Xinyuan Cui, Binbin Li, Yi Lin, Ning Xu, Meng Li, Yi Xu, Xiaoxin Qi, Xiaojuan Wang, Zhongrui Zhang, Xumeng Shang, Dashan Wang, Han Liu, Qi Cheng, Kwang-Ting Liu, Ming Department of Electrical and Electronic Engineering the University of Hong Kong Hong Kong ACCESS – AI Chip Center for Emerging Smart Systems InnoHK Centers Hong Kong Science Park Hong Kong Laboratory of Microelectronic Devices and Integrated Technology Institute of Microelectronics Chinese Academy of Sciences Beijing100029 China State Key Laboratory of Integrated Chips and Systems Frontier Institute of Chip and System Fudan University Shanghai200433 China Institute of the Mind the University of Hong Kong Hong Kong University of Chinese Academy of Sciences Beijing100049 China School of Integrated Circuits Hubei Key Laboratory for Advanced Memories Huazhong University of Science and Technology Wuhan430074 China Department of Electronic and Computer Engineering the Hong Kong University of Science and Technology Hong Kong

Human brains image complicated scenes when reading a novel. Replicating this imagination is one of the ultimate goals of AI-Generated Content (AIGC). However, current AIGC methods, such as score-based diffusion, are still deficient in terms of rapidity and efficiency. This deficiency is rooted in the difference between the brain and digital computers. Digital computers have physically separated storage and processing units, resulting in frequent data transfers during iterative calculations, incurring large time and energy overheads. This issue is further intensified by the conversion of inherently continuous and analog generation dynamics, which can be formulated by neural differential equations, into discrete and digital operations. Inspired by the brain, we propose a time-continuous and analog in-memory neural differential equation solver for score-based diffusion, employing emerging resistive memory. The integration of storage and computation within resistive memory synapses surmount the von Neumann bottleneck, benefiting the generative speed and energy efficiency. The closed-loop feedback integrator is time-continuous, analog, and compact, physically implementing an infinite-depth neural network. Moreover, the software-hardware co-design is intrinsically robust to analog noise. We experimentally validate our solution with 180 nm resistive memory in-memory computing macros. Demonstrating equivalent generative quality to the software baseline, our system achieved remarkable enhancements in generative speed for both unconditional and conditional generation tasks, by factors of 64.8 and 156.5, respectively. Moreover, it accomplished reductions in energy consumption by factors of 5.2 and 4.1. Our approach heralds a new horizon for hardware solutions in edge computing for generative AI applications. © 2024, CC BY-NC-SA.

关键词： Diffusion