Regenerative memory in time-delayed neuromorphic photonic resonators.
Ontology highlight
ABSTRACT:
SUBMITTER: Romeira B
PROVIDER: S-EPMC4726037 | biostudies-other | 2016 Jan
REPOSITORIES: biostudies-other
Similar Datasets
Project description:Photonic systems for high-performance information processing have attracted renewed interest. Neuromorphic silicon photonics has the potential to integrate processing functions that vastly exceed the capabilities of electronics. We report first observations of a recurrent silicon photonic neural network, in which connections are configured by microring weight banks. A mathematical isomorphism between the silicon photonic circuit and a continuous neural network model is demonstrated through dynamical bifurcation analysis. Exploiting this isomorphism, a simulated 24-node silicon photonic neural network is programmed using "neural compiler" to solve a differential system emulation task. A 294-fold acceleration against a conventional benchmark is predicted. We also propose and derive power consumption analysis for modulator-class neurons that, as opposed to laser-class neurons, are compatible with silicon photonic platforms. At increased scale, Neuromorphic silicon photonics could access new regimes of ultrafast information processing for radio, control, and scientific computing.
| S-EPMC5547135 | biostudies-literature
Project description:Viruses represent a continual threat to humans through a number of mechanisms, which include disease, bioterrorism, and destruction of both plant and animal food resources. Many contemporary techniques used for the detection of viruses and viral infections suffer from limitations such as the need for extensive sample preparation or the lengthy window between infection and measurable immune response, for serological methods. In order to develop a method that is fast, cost-effective, and features reduced sample preparation compared to many other virus detection methods, we report the application of silicon photonic microring resonators for the direct, label-free detection of intact viruses in both purified samples as well as in a complex, real-world analytical matrix. As a model system, we demonstrate the quantitative detection of Bean pod mottle virus, a pathogen of great agricultural importance, with a limit of detection of 10 ng/mL. By simply grinding a small amount of leaf sample in buffer with a mortar and pestle, infected leaves can be identified over a healthy control with a total analysis time of less than 45 min. Given the inherent scalability and multiplexing capability of the semiconductor-based technology, we feel that silicon photonic microring resonators are well-positioned as a promising analytical tool for a number of viral detection applications.
| S-EPMC3729447 | biostudies-literature
Project description:Scalable, high-capacity, and low-power computing architecture is the primary assurance for increasingly manifold and large-scale machine learning tasks. Traditional electronic artificial agents by conventional power-hungry processors have faced the issues of energy and scaling walls, hindering them from the sustainable performance improvement and iterative multi-task learning. Referring to another modality of light, photonic computing has been progressively applied in high-efficient neuromorphic systems. Here, we innovate a reconfigurable lifelong-learning optical neural network (L2ONN), for highly-integrated tens-of-task machine intelligence with elaborated algorithm-hardware co-design. Benefiting from the inherent sparsity and parallelism in massive photonic connections, L2ONN learns each single task by adaptively activating sparse photonic neuron connections in the coherent light field, while incrementally acquiring expertise on various tasks by gradually enlarging the activation. The multi-task optical features are parallelly processed by multi-spectrum representations allocated with different wavelengths. Extensive evaluations on free-space and on-chip architectures confirm that for the first time, L2ONN avoided the catastrophic forgetting issue of photonic computing, owning versatile skills on challenging tens-of-tasks (vision classification, voice recognition, medical diagnosis, etc.) with a single model. Particularly, L2ONN achieves more than an order of magnitude higher efficiency than the representative electronic artificial neural networks, and 14× larger capacity than existing optical neural networks while maintaining competitive performance on each individual task. The proposed photonic neuromorphic architecture points out a new form of lifelong learning scheme, permitting terminal/edge AI systems with light-speed efficiency and unprecedented scalability.
| S-EPMC10894876 | biostudies-literature
Project description:Non-Markovian effect is found to be able to decrease the quantum speed limit (QSL) time, and hence to enhance the intrinsic speed of quantum evolution. Although a reservoir with larger degree of non-Markovianity may seem like it should cause smaller QSL times, this seemingly intuitive thinking may not always be true. We illustrate this by investigating the QSL time of a qubit that is coupled to a two-band photonic-band-gap (PBG) environment. We show how the QSL time is influenced by the coherent property of the reservoir and the band-gap width. In particular, we find that the decrease of the QSL time is not attributed to the increasing non-Markovianity, while the memory time of the environment can be seen as an essential reflection to the QSL time. So, the QSL time provides a further insight and sharper identification of memory time in a PBG environment. We also discuss a feasible experimental realization of our prediction.
| S-EPMC5180220 | biostudies-literature
Project description:This report discusses the electrical characteristics of two-terminal synaptic memory devices capable of demonstrating an analog change in conductance in response to the varying amplitude and pulse-width of the applied signal. The devices are based on Mn doped HfO₂ material. The mechanism behind reconfiguration was studied and a unified model is presented to explain the underlying device physics. The model was then utilized to show the application of these devices in speech recognition. A comparison between a 20 nm × 20 nm sized synaptic memory device with that of a state-of-the-art VLSI SRAM synapse showed ~10× reduction in area and >10(6) times reduction in the power consumption per learning cycle.
| S-EPMC4061545 | biostudies-other
Project description:A major characteristic of spiking neural networks (SNNs) over conventional artificial neural networks (ANNs) is their ability to spike, enabling them to use spike timing for coding and efficient computing. In this paper, we assess if neuromorphic datasets recorded from static images are able to evaluate the ability of SNNs to use spike timings in their calculations. We have analyzed N-MNIST, N-Caltech101 and DvsGesture along these lines, but focus our study on N-MNIST. First we evaluate if additional information is encoded in the time domain in a neuromorphic dataset. We show that an ANN trained with backpropagation on frame-based versions of N-MNIST and N-Caltech101 images achieve 99.23 and 78.01% accuracy. These are comparable to the state of the art-showing that an algorithm that purely works on spatial data can classify these datasets. Second we compare N-MNIST and DvsGesture on two STDP algorithms, RD-STDP, that can classify only spatial data, and STDP-tempotron that classifies spatiotemporal data. We demonstrate that RD-STDP performs very well on N-MNIST, while STDP-tempotron performs better on DvsGesture. Since DvsGesture has a temporal dimension, it requires STDP-tempotron, while N-MNIST can be adequately classified by an algorithm that works on spatial data alone. This shows that precise spike timings are not important in N-MNIST. N-MNIST does not, therefore, highlight the ability of SNNs to classify temporal data. The conclusions of this paper open the question-what dataset can evaluate SNN ability to classify temporal data?
| S-EPMC8027306 | biostudies-literature
Project description:We describe an approach for multiplexed microRNA analysis using silicon photonic microring resonators to detect cDNA reverse transcription products via a subsequent enzymatic signal enhancement strategy. Key to this method is a modified stem loop primer that facilitates downstream signal amplification via enzymatic turnover and improves the sensor signal 20-fold when compared to traditional stem loop primers. This approach facilitates targeted microRNA quantification in only 2.5 h and without requiring target amplification via the polymerase chain reaction (PCR). Primers for 7 miRNA targets were orthogonally designed to avoid cross-hybridization between capture probes. This approach was applied to the detection of total RNA from human tissues and found to display differential expression profiles consistent with literature precedent. This development holds promise as an alternative to single-plex RT-qPCR methods and more expensive RNA-seq by offering a cost-effective method to analyze targeted miRNA panels in emerging diagnostic applications.
| S-EPMC5812356 | biostudies-literature
Project description:Cholesteric liquid crystals (CLCs) have a photonic bandgap due to the periodic change of refractive index along their helical axes. The CLCs containing optical gain have served as band-edge lasing resonators. In particular, CLCs in a granular format provide omnidirectional lasing, which are promising as a point light source. However, there is no platform that simultaneously achieves high stability in air and wavelength tunability. We encapsulate CLCs with double shells to design a capsule-type laser resonator. The fluidic CLCs are fully enclosed by an aqueous inner shell that promotes the planar alignment of LC molecules along the interface. The outer shell made of silicone elastomer protects the CLC core and the inner shell from the surroundings. Therefore, the helical axes of the CLCs are radially oriented within the capsules, which provide a stable omnidirectional lasing in the air. At the same time, the fluidic CLCs enable the fine-tuning of lasing wavelength with temperature. The capsules retain their double-shell structure during the dynamic deformation. Therefore, the CLCs in the core maintain the planar alignment along the deformed interface, and a lasing direction can be varied from omnidirectional to bi- or multidirectional, depending on the shape of deformed capsules.
| S-EPMC6014715 | biostudies-literature
Project description:Despite the recent emergence of microcavity resonators as label-free biological and chemical sensors, practical applications still require simple and robust methods to impart chemical selectivity and reduce the cost of fabrication. We introduce the use of hydrocarbon-in-fluorocarbon-in-water (HC/FC/W) double emulsions as a liquid top cladding that expands the versatility of optical resonators as chemical sensors. The all-liquid complex emulsions are tunable droplets that undergo dynamic and reversible morphological transformations in response to a change in the chemical environment (e.g., exposure to targeted analytes). This chemical-morphological coupling drastically modifies the effective refractive index, allowing the complex emulsions to act as a chemical transducer and signal amplifier. We detect this large change in the refractive index by tracking the shift of the enveloped resonant spectrum of a silicon nitride (Si3N4) racetrack resonator-based sensor, which correlates well with a change in the morphology of the complex droplets. This combination of soft materials (dynamic complex emulsions) and hard materials (on-chip resonators) provides a unique platform for liquid-phase, real-time, and continuous detection of chemicals and biomolecules for miniaturized and remote, environmental, medical, and wearable sensing applications.
| S-EPMC7384970 | biostudies-literature
Project description:A label-free biosensing method for the sensitive detection and identification of bacterial transfer-messenger RNA (tmRNA) is presented employing arrays of silicon photonic microring resonators. Species specific tmRNA molecules are targeted by complementary DNA capture probes that are covalently attached to the sensor surface. Specific hybridization is monitored in near real-time by observing the resonance wavelength shift of each individual microring. The sensitivity of the biosensing platform allowed for detection down to 53 fmol of Streptococcus pneumoniae tmRNA, equivalent to approximately 3.16×10(7) CFU of bacteria. The simplicity and scalability of this biosensing approach makes it a promising tool for the rapid identification of different bacteria via tmRNA profiling.
| S-EPMC3367093 | biostudies-literature