Abstract: A CMOS-based resistive processing unit (RPU) and method for a neural network. The RPU includes a capacitor device configured to store a charge representing a weight value associated with a neural network circuit operation. A current source Field Effect Transistor (FET) device is operatively connected to the capacitor device for increasing a charge on the capacitor. A current sink FET device operatively connected to the capacitor device is configured for decreasing the stored capacitor charge. An analog weight update circuit receives one or more update signals generated in conjunction with the neural network circuit operation, the analog weight update circuit controlling the current source FET device and the current sink FET device to provide either a determined amount of current to increase the stored charge on the capacitor device, or sink a determined amount of current to decrease the stored charge on the capacitor device.

Abstract: A method for forming a floating gate memory cell includes: forming an active region on a semiconductor substrate; forming a gate stack on the active region, the gate stack including a first gate layer defining a floating gate of the memory cell structure, a dielectric layer formed on the first gate layer, and a second gate layer defining a control gate of the memory cell structure formed on the dielectric layer; forming first and second source/drain regions in the active region, self-aligned with the gate stack; forming an erase/injection gate on at least a portion of the dielectric layer and spaced laterally from the control gate, the erase/injection gate being proximate to and above the floating gate; and forming multiple contacts providing electrical connection with the first and second source/drain regions, the control gate and the erase/injection gate.

Abstract: A CMOS-based resistive processing unit (RPU) for a neural network. The RPU includes a capacitor device configured to store a charge representing a weight value associated with a neural network circuit operation. A current source Field Effect Transistor (FET) device is operatively connected to the capacitor device for increasing a charge on the capacitor. A current sink FET device operatively connected to the capacitor device is configured for decreasing the stored capacitor charge. An analog weight update circuit receives one or more update signals generated in conjunction with the neural network circuit operation, the analog weight update circuit controlling the current source FET device and the current sink FET device to provide either a determined amount of current to increase the stored charge on the capacitor device, or sink a determined amount of current to decrease the stored charge on the capacitor device.

Abstract: A memory device including a plurality of memory unit cells arranged in a crossbar configuration for a neural network is provided. Each of the memory unit cells includes a readout transistor, a charging transistor, a discharging transistor, and a metal-insulator-metal (MIM) capacitor connected to one of source/drain regions of each of the charging transistor and the discharging transistor and a functional gate of the readout transistor for storing analog information.

Abstract: A resistive processing unit (RPU) circuit for use in a neural network application includes at least one floating gate storage device, the floating gate storage device including a floating gate, a control gate and an inject/erase gate. The RPU circuit further includes a feedback circuit connected with the floating gate storage device. The feedback circuit is configured to maintain a substantially constant floating gate potential of the floating gate storage device during an update mode of operation of the RPU circuit, and is disabled during a readout mode of operation of the RPU circuit.

Abstract: A metal-insulator-metal (MIM) capacitor structure includes source and drain regions formed within a semiconductor substrate, a first conducting layer formed over the source and drain regions, and a dielectric layer formed over the first conducting layer. The MIM capacitor structure further includes a second conducting layer formed over the dielectric layer, and a sidewall dielectric formed adjacent the first conducting layer and the dielectric layer. An electric field is created indirectly through the sidewall dielectric to an adjacent field effect transistor (FET) channel in the semiconductor substrate.

Abstract: A memory device including a plurality of memory unit cells arranged in a crossbar configuration for a neural network is provided. Each of the memory unit cells includes a readout transistor, a charging transistor, a discharging transistor, and a metal-insulator-metal (MIM) capacitor connected to one of source/drain regions of each of the charging transistor and the discharging transistor and a functional gate of the readout transistor for storing analog information.

Abstract: A pickup apparatus includes a digital signal processor, at least one microphone, multiple amplifiers, and multiple analog-to-digital converters (ADCs), where each microphone is configured to receive an audio signal, and send the audio signal to an amplifier corresponding to the microphone, each amplifier is configured to amplify the received audio signal, and send the amplified audio signal to an analog-to-digital converter corresponding to the amplifier, each analog-to-digital converter is configured to convert the received audio signal into a digital signal, and send the digital signal to the digital signal processor, and the digital signal processor is configured to receive multiple digital signals sent by the multiple analog-to-digital converters, determine an unclipped digital signal from the multiple received digital signals, and determine an output signal according to a gain value corresponding to the unclipped digital signal.

Abstract: An apparatus for correcting the shape of a reticle, including: suckers; a sucker mounting frame; and a pneumatic control system, wherein the sucker mounting frame is disposed above a reticle stage, the suckers are mounted on a bottom of the sucker mounting frame in a manner of being spaced apart from one another; and the pneumatic control system is configured to control the suckers to retain the reticle by suction or to release the reticle. The suckers are so arranged that they do not block trapezoidal exposure fields of view on the reticle, thereby allowing a light beam to be fully (100%) incident on the reticle, which results in improved exposure efficiency. The apparatus is mounted separately from a photolithography tool by which it is employed, enabling a lower weight load on the reticle. Additionally, the apparatus is not in contact with any component disposed in the photolithography tool and will not introduce additional external vibration to the tool.

Abstract: A method of scheduling services is disclosed. The method may include: acquiring scheduling related information; determining a scheduling strategy according to the scheduling related information; and sending the scheduling strategy to at least one of an order initiator or an order recipient, wherein the scheduling related information includes but is not limited to historical order information, real-time order information, historical weather information, real-time weather information, future weather information, traffic information, historical information of a service provider, real-time information of a service provider, and vehicle information collected by a vehicle-mounted diagnosis system. The scheduling strategy may include, but is not limit to a supply and demand density push strategy, a hotspot characteristic push strategy, a statistical characteristic push strategy, an order push strategy, an order adjustment strategy, or a prompt information push strategy.

Abstract: A system for pushing orders is disclosed. The system may include: a computer readable storage medium configured to store an executable module, comprising a service requester interface configured to: acquire at least one first order and information related to the first order, comprising geographic and time information related to the first order; and acquire at least one second order and information related to the second order, comprising geographic and time information related to the second order; a processing module configured to filter the second order based on the information related to the first order and the second order to generate at least one filtered second order; a service provider interface configured to send the filtered second order; and a processor configured to implement the executable module. Also disclosed herein are a method of pushing orders and a method of using a system for pushing orders.

Abstract: A pickup apparatus includes a digital signal processor, at least one microphone, multiple amplifiers, and multiple analog-to-digital converters (ADCs), where each microphone is configured to receive an audio signal, and send the audio signal to an amplifier corresponding to the microphone, each amplifier is configured to amplify the received audio signal, and send the amplified audio signal to an analog-to-digital converter corresponding to the amplifier, each analog-to-digital converter is configured to convert the received audio signal into a digital signal, and send the digital signal to the digital signal processor, and the digital signal processor is configured to receive multiple digital signals sent by the multiple analog-to-digital converters, determine an unclipped digital signal from the multiple received digital signals, and determine an output signal according to a gain value corresponding to the unclipped digital signal.

Abstract: An apparatus for correcting the shape of a reticle, including: suckers; a sucker mounting frame; and a pneumatic control system, wherein the sucker mounting frame is disposed above a reticle stage, the suckers are mounted on a bottom of the sucker mounting frame in a manner of being spaced apart from one another; and the pneumatic control system is configured to control the suckers to retain the reticle by suction or to release the reticle. The suckers are so arranged that they do not block trapezoidal exposure fields of view on the reticle, thereby allowing a light beam to be fully (100%) incident on the reticle, which results in improved exposure efficiency. The apparatus is mounted separately from a photolithography tool by which it is employed, enabling a lower weight load on the reticle. Additionally, the apparatus is not in contact with any component disposed in the photolithography tool and will not introduce additional external vibration to the tool.

Abstract: The present invention relates to a tail for improving anti-bird strike performance of an aircraft. A leading edge reinforcement having a shape of an isosceles triangle is located inside a tail leading edge. The leading edge reinforcement is spanwisely fixed in sections between respective spans formed by the wing rib inside the tail leading edge along the tail of the aircraft. An apex angle of the leading edge reinforcement is the same as an apex angle or arc transition of the tail leading edge skin. The leading edge reinforcement is fixedly connected with the small front beam by a leading edge reinforcement fixed surface. The present invention additionally installs a leading edge reinforcement in the original tail of the aircraft.

Abstract: The present invention relates to call voice signal processing, and particularly, to apparatus and method for echo control in parameter domain. The apparatus includes: an echo detection module configured to extract parameter domain characteristic parameters of a far-end output signal and a near-end input signal, respectively, through parameter domain decoding, and detect whether a near-end input signal frame is an echo frame according to the parameter domain characteristic parameters; and a transition module configured to re-quantize the fixed codebook gain, the adaptive codebook gain and an LSF prediction error in the near-end input signal frame when the echo detection module detects the near-end input signal frame as a non-echo frame, and replace the original fixed codebook gain, the original adaptive codebook gain and the original LSF in the near-end input signal frame.

Abstract: The present invention relates to call voice signal processing, and particularly, to apparatus and method for echo control in parameter domain. The apparatus includes: an echo detection module configured to extract parameter domain characteristic parameters of a far-end output signal and a near-end input signal, respectively, through parameter domain decoding, and detect whether a near-end input signal frame is an echo frame according to the parameter domain characteristic parameters; and a transition module configured to re-quantize the fixed codebook gain, the adaptive codebook gain and an LSF prediction error in the near-end input signal frame when the echo detection module detects the near-end input signal frame as a non-echo frame, and replace the original fixed codebook gain, the original adaptive codebook gain and the original LSF in the near-end input signal frame.

Abstract: The present invention relates to a catalyst composition for ethylene oligomerization and the use thereof. Such catalyst composition includes chromium compound, ligand containing P and N, activator and accelerator; wherein the chromium compound is selected from the group consisting of acetyl acetone chromium, THF-chromium chloride and Cr(2-ethylhecanoate)3; general formula of the ligand containing P and N is shown as: in which R1, R2, R3 and R4 are phenyl, benzyl, or naphthyl. R5 is isopropyl, butyl, cyclopropyl, cyclopentyl, cyclohexyl or fluorenyl; the activatior is methyl aluminoxane, ethyl aluminoxane, propyl aluminoxane and/or butyl aluminoxane; the accelerator is selected from the group consisting of 1,1,2,2,-tetrachloroethane, 1,1,2,2-tetrabromoethane, 1,1,2,2-tetrafluoroethane, and compounds having a formula of X1R6X2, in which X1 and X2 are F, Cl, Br, I or alkoxyl, R6 is alkylene or arylene group; the molar ratio of chromium compound, ligand containing P and N, activator and accelerator is 1:0.

Abstract: A device and a method for frame lost concealment are disclosed. A pitch period of a current lost frame is obtained on the basis of a pitch period of the last good frame before the current lost frame. An excitation signal of the current lost frame is recovered on the basis of the pitch period of the current lost frame and an excitation signal of the last good frame before the lost frame. Thereby, the hearing contrast of a receiver is reduced, and the quality of speech is improved. Further, in the present invention, a pitch period of continual lost frames is adjusted on the basis of the change trend of the pitch period of the last good frame before the lost frame. Therefore, a buzz effect produced by the continual lost frames is avoided, and the quality of speech is further improved.

Abstract: A device and a method for frame lost concealment are disclosed. A pitch period of a current lost frame is obtained on the basis of a pitch period of the last good frame before the current lost frame. An excitation signal of the current lost frame is recovered on the basis of the pitch period of the current lost frame and an excitation signal of the last good frame before the lost frame. Thereby, the hearing contrast of a receiver is reduced, and the quality of speech is improved. Further, in the present invention, a pitch period of continual lost frames is adjusted on the basis of the change trend of the pitch period of the last good frame before the lost frame. Therefore, a buzz effect produced by the continual lost frames is avoided, and the quality of speech is further improved.

Abstract: The present invention relates to a catalyst composition for ethylene oligomerization and the use thereof. Such catalyst composition includes chromium compound, ligand containing P and N, activator and accelerator; wherein the chromium compound is selected from the group consisting of acetyl acetone chromium, THF-chromium chloride and/or Cr(2-ethylhecanoate)3; general formula of the ligand containing P and N is shown as: in which R1, R2, R3 and R4 are phenyl, benzyl, or naphthyl. R5 is isopropyl, butyl, cyclopropyl, cyclopentyl, cyclohexyl or fluorenyl; the activatior is methyl aluminoxane, ethyl aluminoxane, propyl aluminoxane and/or butyl aluminoxane; general formula of the accelerator is X1R6X2, in which X1 and X2 are F, Cl, Br, I or alkoxyl, R6 is alkyl or aryl; the molar ratio of a, b, c and d is 1:0.5˜10:50˜3000:0.5˜10.