中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

22 records (took 0.02 seconds)

    All-Photonic Artificial Neural Network Processor Via Nonlinear Optics
    1.
    发明公开
    All-Photonic Artificial Neural Network Processor Via Nonlinear Optics 审中-公开

    公开(公告)号：US20230351168A1

    公开(公告)日：2023-11-02

    申请号：US18310697

    申请日：2023-05-02

    Applicant: Massachusetts Institute of Technology

    Inventor： Jasvith Raj Basani ,  Mikkel HEUCK ,  Dirk Robert ENGLUND ,  Stefan Ivanov Krastanov

    IPC: G06N3/067 ,  G06N3/048

    CPC classification number: G06N3/067 ,  G06N3/048

    Abstract: An all-photonic computational accelerator encodes information in the amplitudes of frequency modes stored in a ring resonator. Nonlinear optical processes enable interaction among these modes. Both the matrix multiplication and element-wise activation functions on these modes (the artificial neurons) occur through coherent processes, enabling the representation of negative and complex numbers without digital electronics. This accelerator has a lower hardware footprint than electronic and optical accelerators, as the matrix multiplication happens in a single multimode resonator on chip. Our architecture provides a unitary, reversible mode of computation, enabling on-chip analog Hamiltonian-echo backpropagation for gradient descent and other self-learning tasks. Moreover, the computational speed increases with the power of the pumps to arbitrarily high rates, as long as the circuitry can sustain the higher optical power.

    APPARATUS, SYSTEMS, AND METHODS FOR NONBLOCKING OPTICAL SWITCHING
    2.
    发明申请
    APPARATUS, SYSTEMS, AND METHODS FOR NONBLOCKING OPTICAL SWITCHING 审中-公开

    公开(公告)号：US20200284989A1

    公开(公告)日：2020-09-10

    申请号：US16827795

    申请日：2020-03-24

    Applicant: Massachusetts Institute of Technology

    Inventor： Gregory R. STEINBRECHER ,  Dirk Robert ENGLUND

    IPC: G02B6/35 ,  G02B6/293 ,  H04Q11/00

    Abstract: A method of nonblocking optical switching includes guiding a first optical beam from a first input to a first output via a first path through an optical switching fabric. The first path traverses a phase shifter disposed between a pair of cascaded Mach-Zehnder interferometers. The method also includes receiving a second optical beam for a second path intersecting with the first path through the optical switching fabric. The method also includes moving the first optical beam from the first path to a third path connecting the first input to the first output without intersecting the second path. The method also includes shifting a phase of the first optical beam, with the phase shifter, while moving the first optical beam from the first path to the third path to prevent the first optical beam from interfering with the second optical beam.

    ALL-RESONANT ACTUATION OF PHOTONIC INTEGRATED CIRCUITS
    3.
    发明公开
    ALL-RESONANT ACTUATION OF PHOTONIC INTEGRATED CIRCUITS 审中-公开

    公开(公告)号：US20240118537A1

    公开(公告)日：2024-04-11

    申请号：US18480981

    申请日：2023-10-04

    Applicant: The MITRE Corporation ,  Sandia National Laboratories ,  MIT - Massachusetts Institute of Technology

    Inventor： Mark DONG ,  Gerald Neal GILBERT ,  Matthew Scott EICHENFIELD ,  Dirk Robert ENGLUND

    IPC: G02B26/10

    CPC classification number: G02B26/103

    Abstract: Provided herein is a photonic integrated circuit and methods for controlling a photonic integrated circuit that can utilize the resonant frequency of one or more components of the photonic integrated circuit to enhance the response of the circuit. At least one component of the photonic integrated circuit can be driven by an electrical signal whose frequency is substantially equal to the mechanical resonance frequency of the component such that the response of the optical component is increased. The component of the photonic integrated circuit can include a phase shifter that can impart a phase shift on a received optical signal. By driving the phase shifter with an electrical signal that is equal to the mechanical resonance frequency of the optical phase shifter, less power can be required to impart a desired phase shift on a received optical signal. The optical components can be implemented using piezoelectric cantilevers.

    Low-Power Edge Computing with Optical Neural Networks via WDM Weight Broadcasting
    5.
    发明公开
    Low-Power Edge Computing with Optical Neural Networks via WDM Weight Broadcasting 审中-公开

    公开(公告)号：US20230274156A1

    公开(公告)日：2023-08-31

    申请号：US18247129

    申请日：2021-07-29

    Applicant: Ryan HAMERLY ,  Dirk Robert ENGLUND ,  Massachusetts Institute of Technology

    Inventor： Ryan HAMERLY ,  Dirk Robert ENGLUND

    IPC: G06N3/098 ,  G06N5/04

    CPC classification number: G06N3/098 ,  G06N5/04

    Abstract: NetCast is an optical neural network architecture that circumvents constraints on deep neural network (DNN) inference at the edge. Many DNNs have weight matrices that are too large to run on edge processors, leading to limitations on DNN inference at the edge or bandwidth bottlenecks between the edge and server that hosts the DNN. With NetCast, a weight server stores the DNN weight matrix in local memory, modulates the weights onto different spectral channels of an optical carrier, and distributes the weights to one or more clients via optical links. Each client stores the activations, or layer inputs, for the DNN and computes the matrix-vector product of those activations with the weights from the weight server in the optical domain. This multiplication can be performed coherently by interfering the spectrally multiplexed weights with spectrally multiplexed activations or incoherently by modulating the weight signal from the weight server with the activations.

    Absorption-Based Diamond Spin Microscopy on a Plasmonic Quantum Metasurface
    6.
    发明申请
    Absorption-Based Diamond Spin Microscopy on a Plasmonic Quantum Metasurface 有权

    公开(公告)号：US20220082639A1

    公开(公告)日：2022-03-17

    申请号：US17376234

    申请日：2021-07-15

    Applicant: Massachusetts Institute of Technology

    Inventor： Laura KIM ,  Hyeongrak CHOI ,  Matthew Edwin TRUSHEIM ,  Dirk Robert ENGLUND

    IPC: G01R33/032 ,  G01R29/08

    Abstract: Nitrogen vacancy (NV) centers in diamond combine exceptional sensitivity with nanoscale spatial resolution by optically detected magnetic resonance (ODMR). Infrared (IR)-absorption-based readout of the NV singlet state transition can increase ODMR contrast and collection efficiency. Here, a resonant diamond metallodielectric metasurface amplifies IR absorption by concentrating the optical field near the diamond surface. This plasmonic quantum sensing metasurface (PQSM) supports plasmonic surface lattice resonances and balances field localization and sensing volume to optimize spin readout sensitivity. Combined electromagnetic and rate-equation modeling suggests a near-spin-projection-noise-limited sensitivity below 1 nT Hz−1/2 per m2 of sensing area using numbers for contemporary NV diamond samples and fabrication techniques. The PQSM enables microscopic ODMR sensing with IR readout near the spin-projection-noise-limited sensitivity, making it appealing for imaging through scattering tissues and spatially resolved chemical NMR detection.

    VCSEL-based Coherent Scalable Deep Learning
    7.
    发明申请
    VCSEL-based Coherent Scalable Deep Learning 有权

    公开(公告)号：US20250111218A1

    公开(公告)日：2025-04-03

    申请号：US18863245

    申请日：2023-05-15

    Applicant: Massachusetts Institute of Technology

    Inventor： Zaijun Chen ,  Ryan HAMERLY ,  Dirk Robert ENGLUND

    IPC: G06N3/067

    Abstract: The exponential growth in deep learning models is challenging existing computing hardware. Optical neural networks (ONNs) accelerate machine learning tasks with potentially ultrahigh bandwidth and nearly no loss in data movement. Scaling up ONNs involves improving scalability, energy efficiency, compute density, and inline nonlinearity. However, realizing all these criteria remains an unsolved challenge. Here, we demonstrate a three-dimensional spatial time-multiplexed ONN architecture based on dense arrays of microscale vertical cavity surface emitting lasers (VCSELs). The VCSELs, coherently injection-locked to a leader laser, operate at gigahertz data rates with a 7T-phase-shift voltage on the 10-millivolt level. Optical nonlinearity is incorporated into the ONN with no added energy cost using coherent detection of optical interference between VCSELs.

Patent Agency Ranking