43 results about "Neuromorphic circuits" patented technology

Embodiments of the present invention are directed to neuromorphic circuits containing two or more internal neuron computational units. Each internal neuron computational unit includes a synchronization-signal input for receiving a synchronizing signal, at least one input for receiving input signals, and at least one output for transmitting an output signal. A memristive synapse connects an output signal line carrying output signals from a first set of one or more internal neurons to an input signal line that carries signals to a second set of one or more internal neurons.

A neuromorphic circuit performs functions representative of spiking timing dependent plasticity of a synapse.

A synapse for a spike timing dependent (STDP) function cell includes a memory device having a variable resistance, such as a memristor, and a transistor connected to the memory device. A channel of the memory device is connected in series with a channel of the transistor.

An analog neuromorphic circuit is disclosed having resistive memories that provide a resistance to an input voltage signal as the input voltage signal propagates through the resistive memories generating a first output voltage signal and to provide a resistance to a first error signal that propagates through the resistive memories generating a second output voltage signal. A comparator generates the first error signal that is representative of a difference between the first output voltage signal and the desired output signal and generates the first error signal so that the first error signal propagates back through the plurality of resistive memories. A resistance adjuster adjusts a resistance value associated with each resistive memory based on the first error signal and the second output voltage signal to decrease the difference between the first output voltage signal and the desired output signal.

The invention relates to a low-power-consumption neural synapse thin film transistor and a preparation method thereof. The transistor structurally comprises a back gate electrode, a gate dielectric layer, a conductive channel and a source-source electrode from bottom to top in sequence. The source-drain electrode is arranged on the upper surface of the gate dielectric layer, and the conductive channels are located on the upper surface and two sides of the source-drain electrode; channels are formed on the upper surface of the source-drain electrode. Compared with the prior art, the transistorhas ultra-low power consumption at a normal focus level; the adjustable memory time from milliseconds to thousands of seconds can be realized at different dielectric layer preparation temperatures; meanwhile, the stability of the transistor is greatly improved due to the use of a full-inorganic material; the flexibility and synaptic performance of the transistor can be used for flexible electronics and large-scale neuromorphic circuit systems.

A memory capacitor based on a field configurable ion-doped polymer is reported. The device can be dynamically and reversibly programmed to analog capacitances with low-voltage (<5 V) pulses. After the device is programmed to a specific value, its capacitance remains nonvolatile. The field configurable capacitance is attributed to the modification of ionic dopant concentrations in the polymer. The ion and dipole concentrations in the ion conductive layer can be modified when the voltage biases applied to the electrodes exceeds a threshold value and can operate as a conventional capacitor when a voltage less than the threshold value is applied. The ion conductive layer will remain at a stable value after the device is modified without applying external voltage. The device has a nonvolatile memory function even when the external voltage is turned off. The memory capacitors may be used for analog memory, nonlinear analog and neuromorphic circuits.

The invention discloses a robot control system and method based on a memristor intersection array. The system involves a sensor group, an input sensing signal modulator, a neuromorphic circuit, an output control signal modulator, an output device, an external supervision module and a training controller; and control of a robot is performed by the neuromorphic circuit, and the main part of the neuromorphic circuit is the memristor intersection array with a fully connected neural network structure. A differential amplification circuit and a multiplexing switch in the neuromorphic circuit are connected to the memristor intersection array, an input signal vector is multiplied with a weight matrix stored in the memristor intersection array, and then one or more analog output signals are obtained through the differential amplification circuit. According to the robot control system and method, the memristor neuromorphic circuit is used for reducing the power consumption of an existing robot control system in performing neural network operation; and memristors are directly used for simulating the weight in a biological neural network, the complexity is effectively reduced, and the responsespeed is improved.

A neuromorphic computing system is provided which comprises: a synapse core; and a pre-synaptic neuron, a first post-synaptic neuron, and a second post-synaptic neuron coupled to the synaptic core, wherein the synapse core is to: receive a request from the pre-synaptic neuron, generate, in response to the request, a first address of the first post-synaptic neuron and a second address of the second post-synaptic neuron, wherein the first address and the second address are not stored in the synapse core prior to receiving the request.

This invention relates to an apparatus, system, and method for computing with neuromorphic circuit architectures that have neurons with interconnected internal state information. The interconnected internal state information allows the neurons to enable or strengthen the input to other neurons. Furthermore, neuron internal state information provides insights on the characteristics of the input data that can be used to enhance the performance of the neuromorphic system. The neuromorphic system can be implemented with an artificial phase-change-based neurons.

Neuromorphic circuits are multi-cell networks configured to imitate the behavior of biological neural networks. A neuromorphic circuit is provided which comprises a network of neurons each identified by a neuron address in the network, each neuron being able to receive and process at least one input signal and then later emit on an output of the neuron a signal representing an event which occurs inside the neuron, and a programmable memory composed of elementary memories each associated with a respective neuron. The elementary memory, which is a memory of post-synaptic addresses and weights, comprises an activation input linked by a conductor to the output of the associated neuron to directly receive an event signal emitted by this neuron without passing through an address encoder or decoder. The post-synaptic addresses extracted from an elementary memory activated by a neuron are applied, with associated synaptic weights, as inputs to the neural network.

A semiconductor device includes a first transistor including a first channel layer of a first conductivity type, a second transistor provided in parallel with the first transistor and including a second channel layer of a second conductivity type, and a third transistor stacked on the first and second transistors. The third transistor may include a gate insulating film including a ferroelectric material. The third transistor may include third channel layer and a gate electrode that are spaced apart from each other in a thickness direction with the gate insulating film therebetween.

A control device of an array including neuromorphic elements that multiply a signal by a weight corresponding to a value of a variable characteristic is provided with a control unit which calculates update amounts of element conductances in a neuromorphic array on the basis of weight update amounts from a learning algorithm, and, after applying a write signal for changing conductances in the neuromorphic array, selects certain elements with reference to a predetermined threshold value and applies an additional write signal.

Provided are a neuromorphic device and a neuromorphic circuit using the neuromorphic device. The neuromorphic device is configured to include a first semiconductor region formed on a substrate in a wall shape or a dumbbell shape; first, second, third, and fourth doped regions sequentially formed in the first semiconductor region; first and second gate insulating film stacks disposed on the respective side surfaces of the second doped region; first and second gate electrodes respectively disposed on the respective side surfaces of the second doped region; the first and second gate electrodes disposed on the respective side surface of the second doped region, the first and second gate electrodes being electrically insulated from the second doped, region by the first and second gate insulating film stacks; and first and second electrodes electrically connected to the first and fourth doped regions, respectively.

The present invention relates to a neuron circuit and a neuromorphic circuit, the neuron circuit includes a pulse generating circuit and a U unit connected to the control terminal of the pulse generating circuit; the pulse generating circuit is used to convert the input current into a pulse sequence , the U unit is used to control the excitation mode and pulse interval of the pulse generated by the pulse generating circuit. The neuron circuit and the neuromorphic circuit provided by the present invention use the U unit to control the pulse generation circuit to achieve different biological characteristics without using any bias voltage. The method realizes different biological characteristics, avoids the sensitivity of the circuit itself to the fluctuation of process, voltage and temperature, and makes the circuit easier to realize; at the same time, there is no normally open path in the circuit, so that the circuit consumes less power.