Top Research Papers on Neuromorphic Computing

Approaches for the development of future at-scale neuromorphic systems based on principles of biointelligence are described, along with potential applications of scalable neuromorphic architectures and the challenges that need to be overcome.

The need and the possibility to conceive a photonic memristor are discussed, a positive outlook on the challenges and opportunities for the ambitious goal of realising the next generation of full-optical neuromorphic hardware is offered.

Inspired by recent progress in optogenetics and visual sensing, light has been increasingly incorporated into synaptic devices, paves the way to optoelectronic synaptic devices with a series of advantages such as wide bandwidth, negligible resistance–capacitance delay and power loss, and global regulation of multiple synaptic devices.

This survey reviews computing models and hardware platforms of existing neuromorphic computing systems, and introduces neuron and synapse models, followed by the discussion on how they will affect hardware design.

There is still a wide variety of challenges that restrict the rapid growth of neuromorphic algorithmic and application development, and opportunities for future development of algorithms and applications on these systems are discussed.

An overview of various neural networks with a focus on building a memristor‐based spike neural network neuromorphic computing system is provided, and an outlook for brain‐like computing is proposed.

Versatile synaptic functions of the nervous system, including paired‐pulse facilitation, short‐term plasticity, long‐term Plasticity, transition from short‐to‐long‐term memory, and learning‐experience behavior, are successfully emulated.

This review of synaptic devices that mimic synaptic functions is discussed by categorizing them into electrically stimulated, optically stimulated, and photoelectric synergetic synaptic devices based on stimulation of electrical and optical signals.

Recent advances in integrated photonic neuromorphic neuromorphic systems are reviewed, current and future challenges are discussed, and the advances in science and technology needed to meet those challenges are outlined.

Research in photonic computing has flourished due to the proliferation of optoelectronic components on photonic integration platforms. Photonic integrated circuits have enabled ultrafast artificial neural networks, providing a framework for a new class of information processing machines. Algorithms running on such hardware have the potential to address the growing demand for machine learning and artificial intelligence in areas such as medical diagnosis, telecommunications, and high-performance and scientific computing. In parallel, the development of neuromorphic electronics has highlighted c...

The recent advances of ONNs based on different approaches to photonic multiplexing are reviewed, and the outlook on key technologies needed to further advance these photonicMultiplexing/hybrid-multiplexing techniques of Onns are presented.

Arrays of memcapacitor devices that work via charge shielding can be used to implement artificial neural networks and could potentially offer an energy efficiency of 29,600 tera-operations per second per watt.

It is shown that lattices of exciton-polariton condensates accomplish neuromorphic computing with outstanding accuracy thanks to their high optical nonlinearity and it is demonstrated that this neural network significantly increases the recognition efficiency compared to the linear classification algorithms on one of the most widely used benchmarks, the MNIST problem.

Recent advances in WDM-based ONNs are reviewed, focusing on methods that use integrated microcombs to implement ONN's, and results for human image processing using an optical convolution accelerator operating at 11 Tera operations per second are presented.

The critical characteristic parameters of memristors affecting the performance of RC systems, such as reservoir size and decay time, are identified and discussed, and the challenges this field faces in reliable and accurate task processing are summarized and forecasted.

The progress, challenges, and opportunities for both volatile and nonvolatile memristor in the level of materials, integration technology, algorithm, and system are highlighted in this review.

The discovery of an alloyed memristor with alloyed conduction channels enables stable and controllable device operation with high switching uniformity and allows the fabrication of large-scale crossbar arrays that feature a high device yield and accurate analogue programming capability.

The electrical current-induced creation, motion, detection and deletion of skyrmions in ferrimagnetic multilayers can be used to mimic the behaviour of biological synapses, providing devices that could be used for neuromorphic computing tasks such as pattern recognition.

These insights uncover computing principles, primitives, and algorithms on different levels of abstraction and call for more research into the basis of neural computation and neuronally inspired computing hardware.

How novel material properties enable complex dynamics and define different orders of complexity in memristor devices and systems are discussed, which enable new computing architectures that offer dramatically greater computing efficiency than conventional computers.

Abstract Recently, 2D ferroelectrics have attracted extensive interest as a competitive platform for implementing future generation functional electronics, including digital memory and brain‐inspired computing circuits. Fulfilling their potential requires achieving the interplay between ferroelectricity and electronic characteristics on the device operation level, which is currently lacking since most studies are focused on the verification of ferroelectricity from different 2D materials. Here, by leveraging the ferroelectricity and semiconducting properties of α‐In 2 Se 3 , ferroelectric semi...

This work focuses on the development of nonvolatile memories and neuromorphic computing systems based on QD thin-film solids and discusses the advantageous traits of QDs for novel and optimized memory techniques in both conventional flash memories and emerging memristors.

It is demonstrated that graphene memristors enable weight assignment based on k-means clustering, which offers greater computing accuracy when compared with uniform weight quantization for vector matrix multiplication, an essential component for any artificial neural network.

Brain-inspired neuromorphic computing emulates the biological functions of the human brain to achieve highly intensive data processing with low power consumption. In particular, spiking neural networks (SNNs) that consist of artificial synapses can process spatiotemporal information while enabling energy-efficient neuromorphic computations. Artificial synapses are a key element of sophisticated neuromorphic hardware, so a significant amount of research has been conducted to develop various materials and device structures. Of these, we assess amorphous InGaZnO (IGZO)-based synaptic transistors ...

This review highlights the film preparation methods and the application advances in memory and neuromorphic electronics of porous crystalline materials, involving MOFs, COFs, HOFs, and zeolites.

A comprehensive review on the status of reliability studies of analog RSMs, the reliability requirements, and evaluation criteria and outlook for future reliability research directions in this field is provided.

Abstract Many in-memory computing frameworks demand electronic devices with specific switching characteristics to achieve the desired level of computational complexity. Existing memristive devices cannot be reconfigured to meet the diverse volatile and non-volatile switching requirements, and hence rely on tailored material designs specific to the targeted application, limiting their universality. “Reconfigurable memristors” that combine both ionic diffusive and drift mechanisms could address these limitations, but they remain elusive. Here we present a reconfigurable halide perovskite nanocry...

An all‐optically controlled analog memristor is realized, with memconductance that is reversibly tunable over a continuous range by varying only the wavelength of the controlling light, indicating its potential applications in AOC spiking neural networks for highly efficient optoelectronic NC.

In this review, internal physical dynamics, preparation technologies, and modulation methods are systemically examined as well as the progress, challenges, and possible solutions are proposed for next generation emerging ABO3-based memristive application in artificial intelligence.

A comprehensive review of synaptic devices based on 2D materials is provided, including the advantages of2D materials and heterostructures, various robust multifunctional 2D synaptic devices, and associated neuromorphic applications.

For the first time, an electret-based synaptic transistor (EST) is presented, which successfully performs synaptic behaviors including excitatory/inhibitory postsynaptic current (EPSC/IPSC), paired-pulse facilitation/depression (PPF/PPD), long-term plasticity (LTP) and high-pass filtering.

This survey reviews results that are obtained to date with Loihi across the major algorithmic domains under study, including deep learning approaches and novel approaches that aim to more directly harness the key features of spike-based neuromorphic hardware.

This paper focuses on non-volatile memory technologies and their applications to bio-inspired neuromorphic computing, enabling spike-based machine intelligence and cross-layer optimization across underlying NVM based hardware and learning algorithms can be exploited for resilience in learning and mitigating hardware inaccuracies.

It is shown that the CMOS and memristive devices are assembled in different neuromorphic learning platforms to perform simple cognitive tasks such as classification of spike rate‐based patterns or handwritten digits.

The transparent and flexible Mott transistor based on electrically‐controlled VO2 metal‐insulator transition is believed to open up alternative approaches to developing highly stable synapses for future flexible neuromorphic systems.

Recent advances in memristive materials and strategies that emulate synaptic functions for neuromorphic computing are highlighted and working principles and characteristics of biological neurons and synapses, which can be mimicked by Memristive devices are presented.

A new perspective and understanding is focused on the discussions of synaptic devices based neuromorphic computing applications in artificial intelligence, which enables high-performance super-parallel computing, so that it overcomes the von Neumann bottleneck.

Recent state-of-the-art developments related to organic small molecules for resistive random-access memory devices has been emphasized.

This work proposes the reconfigurable diffractive processing unit, an optoelectronic fused computing architecture based on the diffraction of light, which can support different neural networks and achieve a high model complexity with millions of neurons.

Energy-efficient artificial photonic synapse is designed based on photo-sensitive BP/CdS heterostructure device, which can be used for high-performance brain-inspired neuromorphic computing and shows great potential in high- performance neuromorphic vision systems.