Industrial Technology Research Institute

Sih-Han Li

Chih-Sheng Lin

Shyh-Shyuan Sheu

Jian-Wei Su

Kuan-Wei Chen

Fu-Cheng Tsai

Yu-Hui Lin

Jie Zhang

Tuo-Hung Hou

Erh-Hao Chen

Ya-Wen Yang

Shih-Chieh Chang

Pei-Ling Tseng

Wen-Pin Lin

Chih-He Lin

Tsai-Kan Chien

Chih-Ping Cheng

Chang-Po Chao

Zhe-Hui Lin

Tai-Hsing Lee

Wen-Yu Chen

Peng-I Mei

An apparatus and a method for neural network computation are provided. The apparatus for neural network computation includes a first neuron circuit and a second neuron circuit. The first neuron circuit is configured to execute a neural network computation of at least one computing layer with a fixed feature pattern in a neural network algorithm. The second neuron circuit is configured to execute the neural network computation of at least one computing layer with an unfixed feature pattern in the neural network algorithm. The performance of the first neuron circuit is greater than that of the second neuron circuit.

Apparatus and method for neural network computation

A switching circuit includes a transmission gate, two base control sub-circuits each including a first transistor and a second transistor, a third transistor, and a fourth transistor. The transmission gate includes two I/O terminals, two gate control terminals, and two base control terminals, and is configured to conduct or not conduct the two I/O terminals according to the voltage of the two gate control terminals. The two base voltage control sub-circuits, the third transistor and the fourth transistor forms a double balance circuit structure and is electrically connected to the transmission gate. The double balance circuit changes the voltage of the base control terminals according to the voltage of the I/O terminals of the transmission gate.

Switch circuit and programmable connection chip

A high-frequency component test device including a test key and a test module is provided. The test key includes a front-level key and a back-level key which are arranged symmetrically and have the same electrical length and characteristic impedance. The test module is used to measure an S parameter of the front-level key and the back-level key connected directly and an S parameter of a structure where a device under test (DUT) is added between the front-level key and the back-level key. The test module performs S parameter calculation in the frequency domain and converts the S parameter into an ABCD parameter matrix, and then obtains an ABCD parameter of a de-embedded DUT using a matrix root-opening operation and an inverse matrix operation.

High-frequency component test device and method thereof

An arrayed switch circuit includes a substrate, signal conductive pads and signal expansion pins. The signal conductive pads are disposed on the substrate at intervals, and the signal conductive pads are arranged to form a signal conductive pad array. Each of the signal conductive pads has a row position and a column position in the signal conductive pad array. A row signal switch is provided between any two adjacent signal conductive pads corresponding to the same row position, and a column signal switch is provided between any two adjacent signal conductive pads corresponding to the same column position. The signal expansion pins are connected to the signal conductive pads located on at least one side of the signal conductive pad array through signal expansion switches respectively.

Arrayed switch circuit, switching element and system chip package structure

A computing in memory (CIM) cell includes a memory cell circuit, a first semiconductor element, a second semiconductor element, a third semiconductor element, and a fourth semiconductor element. A first terminal of the first semiconductor element receives a bias voltage. A control terminal of the first semiconductor element is coupled to a computing word-line. A control terminal of the second semiconductor element is coupled to a first data node in the memory cell circuit. A second terminal of the third semiconductor element is adapted to receive a reference voltage. A control terminal of the third semiconductor element receives an inverted signal of the computing word-line. A first terminal of the fourth semiconductor element is coupled to a first computing bit-line. A second terminal of the fourth semiconductor element is coupled to a second computing bit-line.

Computing in memory cell

A computing in memory (CIM) cell includes a memory cell circuit, a first semiconductor element, a second semiconductor element, and a third semiconductor element. A first terminal of the first semiconductor element is coupled to a first computing bit-line. A control terminal of the first semiconductor element is coupled to a computing word-line. A control terminal of the second semiconductor element is coupled to the memory cell circuit. A first terminal of the second semiconductor element is coupled to a second terminal of the first semiconductor element. A first terminal of the third semiconductor element is coupled to a second terminal of the second semiconductor element. A second terminal of the third semiconductor element is coupled to a second computing bit-line. A control terminal of the third semiconductor element receives a bias voltage.

A memory device for in-memory computation includes data channels, a memory cell array, a maximum accumulated weight generating array, a minimum accumulated weight generating array, a reference generator and a comparator. The data channels are selectively enabled according to data input. The memory cell array generates an accumulated data weight value according to the quantity of enabled data channels, a first resistance and a second resistance. The maximum accumulated weight generating array generates a maximum accumulated weight value according to the quantity of enabled data channels and the first resistance. The minimum accumulated weight generating array generates a minimum accumulated weight value according to the quantity of enabled data channels and the second resistance. The reference generator generates reference value(s) according to the maximum and minimum accumulated weight values. The comparator compares the accumulated data weight value with the reference value(s) to generate a data weight state.

Memory device and data weight state determining method for in-memory computation

A position-encoding device includes a sensing device, a filtering device, a calibrating device and a compensating device. The sensing device senses the motion of a moving device to generate first and second signals. The filtering device filters the first and second signals to generate first and second filtering signal. The calibrating device captures the first and second filtering signals to obtain time and phase information of the first and second filtering signals, performs gain and offset calibration on the first and second filtering signals, and performs a phase calibration on the first and second filtering signals through first, second feedback signals and the time and phase information of the first and second filtering signals to generate first and second calibrating signals. The compensating device compensates for the first and second calibrating signals according to a lookup table, so as to generate first and second position encoding signals.

Position-encoding device and method

A readout circuit for a sensor and a readout method thereof are provided. The readout circuit includes a reference circuit, a compensated circuit, and a signal processing circuit. The reference circuit provides a direct current (DC) signal. The compensated circuit is coupled to the reference circuit. The compensated circuit obtains an analog sensing signal of the sensor, obtains the DC signal from the reference circuit, and provides a compensated signal according to the analog sensing signal and the DC signal. The signal processing circuit is coupled to the compensated circuit. The signal processing circuit processes the compensated signal to convert the compensated signal into a digital sensing signal. The compensated circuit subtracts the DC signal from the analog sensing signal to provide the compensated signal.

Readout circuit for sensor and readout method thereof

A method of manufacturing a sensor device is provided. In the method, sensing electrodes are formed on a substrate, a sensing material layer is formed on the sensing electrodes. The sensing material layer is etched to form a first nanowire sensing region, a second nanowire sensing region and a third nanowire sensing region respectively between every two sensing electrodes of the sensing electrodes. A dielectric layer is formed to cover the first nanowire sensing region, the second nanowire sensing region and the third nanowire sensing region, and the first nanowire sensing region and the third nanowire sensing region are exposed.

Method of manufacturing sensor device

A sensor device and a method of manufacturing the same are provided. The sensor device includes a substrate, a plurality of sensing electrodes, a humidity nanowire sensor, a temperature nanowire sensor, and a gas nanowire sensor. The sensing electrodes are formed on the substrate, and the humidity, the temperature and the gas nanowire sensors are also on the substrate. The humidity nanowire sensor includes an exposed first nanowire sensing region, the temperature nanowire sensor includes a second nanowire sensing region, and the gas nanowire sensor includes a third nanowire sensing region.

Sensor device and method of manufacturing the same

A sensor interface circuit and sensor output adjusting method are provided. The sensor interface circuit includes a processor and a gain control circuit. The processor obtains information of a linear region of a sensor to set a configuration corresponding to the sensor. The gain control circuit is coupled to the processor, performs a return-to-zero operation for a maximum electronic value and a minimum electronic value corresponding to the linear region and performs a full-scale operation for a slope of the linear region according to the maximum input range of an analog-to-digital converter which is a subsequent-stage circuit of the sensor interface circuit.

Sensor interface circuit and sensor output adjusting method

A resistive random-access memory device includes a memory array, a read circuit, a write-back logic circuit and a write-back circuit. The read circuit reads the data stored in a selected memory cell and accordingly generates a first control signal. The write-back logic circuit generates a write-back control signal according to the first control signal and a second control signal. The write-back circuit performs a write-back operation on the selected memory cell according to the write-back control signal and a write-back voltage, so as to change a resistance state of the selected memory cell from a low resistance state to a high resistance state, and generates the second control signal according to the resistance state of the selected memory cell.

Resistive random-access memory devices

First and second paths of the primary-coil of the transformer are located at different sides of the symmetry-line. First terminals of the first and second paths are first and second ports of the primary-coil. Second terminals of the first and second paths are connected to each other. Two partial paths of the first path are connected to each other by TSV. Two partial paths of the second path are connected to each other by TSV. Third and fourth paths of the secondary-coil of the transformer are located on different sides of the symmetry-line. First terminals of the third and fourth paths are first and second ports of the secondary-coil. Second terminals of the third and fourth paths are connected to each other. Two partial paths of the third path are connected to each other by TSV. Two partial paths of the fourth path are connected to each other by TSV.

Three-dimension symmetrical vertical transformer

A varactor is provided. A substrate includes a first surface, a second surface and a first opening and a second opening in the substrate. A conductive material is filling the first and second openings, to form a first through-wafer via (TWV) and a second through-wafer via. A first capacitor is coupled between the first through-wafer via and a first terminal. A second capacitor is coupled between the second through-wafer via and a second terminal. A capacitance of a depletion-region capacitor between the first through-wafer via and the second through-wafer via is determined by a bias voltage applied to the first through-wafer via and the second through-wafer via.

Varactor that applies bias voltage to two through wafer vias to determine capacitance of depletion region capacitor formed between the two through wafer vias

Growing community of inventors

Taipei, Taiwan

Sih-Han Li