587results about "Exclusive-OR circuits" patented technology

A multi-chip package includes a field-programmable-gate-array (FPGA) integrated-circuit (IC) chip configured to perform a logic function based on a truth table, wherein the field-programmable-gate-array (FPGA) integrated-circuit (IC) chip comprises multiple non-volatile memory cells therein configured to store multiple resulting values of the truth table, and a programmable logic block therein configured to select, in accordance with one of the combinations of its inputs, one from the resulting values into its output; and a memory chip coupling to the field-programmable-gate-array (FPGA) integrated-circuit (IC) chip, wherein a data bit width between the field-programmable-gate-array (FPGA) integrated-circuit (IC) chip and the memory chip is greater than or equal to 64.

Various circuit techniques for implementing ultra high speed circuits use current-controlled CMOS (C³MOS) logic fabricated in conventional CMOS process technology. An entire family of logic elements including inverter/buffers, level shifters, NAND, NOR, XOR gates, latches, flip-flops and the like are implemented using C³MOS techniques. Optimum balance between power consumption and speed for each circuit application is achieve by combining high speed C³MOS logic with low power conventional CMOS logic. The combined C³MOS/CMOS logic allows greater integration of circuits such as high speed transceivers used in fiber optic communication systems. The C³MOS structure enables the use of a power supply voltage that may be larger than the voltage required by the CMOS fabrication process, further enhancing the performance of the circuit.

Some embodiments of the invention provide an integrated circuit (IC) with a defect-tolerant neural network. The neural network has one or more redundant neurons in some embodiments. After the IC is manufactured, a defective neuron in the neural network can be detected through a test procedure and then replaced by a redundant neuron (i.e., the redundant neuron can be assigned the operation of the defective neuron). The routing fabric of the neural network can be reconfigured so that it re-routes signals around the discarded, defective neuron. In some embodiments, the reconfigured routing fabric does not provide any signal to or forward any signal from the discarded, defective neuron, and instead provides signals to and forwards signals from the redundant neuron that takes the defective neuron's position in the neural network. In some embodiments that implement a neural network by re-purposing (i.e., reconfiguring) one or more individual neurons to implement neurons of multiple stages of the neural network, the IC discards a defective neuron by removing it from the pool of neurons that it configures to perform the operation(s) of neurons in one or more stages of neurons, and assigning this defective neuron's configuration(s) (i.e., its machine-trained parameter set(s)) to a redundant neuron. In some of these embodiments, the IC would re-route around the defective neuron and route to the redundant neuron, by (1) supplying machine-trained parameters and input signals (e.g., previous stage neuron outputs) to the redundant neuron instead of supplying these parameters and signals to the defective neuron, and (2) storing the output(s) of the redundant neuron instead of storing the output(s) of the defective neuron.

A clock and data recovery system using a distributed variable delay line is provided. The clock and data recovery system can use a delay-locked loop methodology to align a local clock with an incoming data stream. The variable delay line can include a transmission line coupled with a plurality of variable capacitors responsive to a control voltage. The variable delay line can also have a ladder configuration of multiple LC subcircuits each having a variable impedance responsive to a control voltage.

Any of several information processing techniques may be used in various information storage and transmission applications to prevent the occurrence of certain "forbidden" bit patterns. According to an encoding technique, a reversible coding process is used to generate an encoded representation of an information stream that cannot contain any forbidden data patterns. This may be accomplished by partitioning the information stream into segments and encoding each segment according to a respective encoding key that is selected such that the results of the coding process cannot contain a forbidden data pattern. According to one substitution technique, all occurrences of forbidden data patterns are replaced with permissible data patterns that do not otherwise occur in the information stream. This may be accomplished by partitioning the information stream into segments, identifying an unused data pattern in a respective segment, and carrying out the replacement of all occurrences of the forbidden data pattern in that segment. According to another substitution technique, all occurrences of a forbidden data pattern are replaced by any permissible data pattern. This may be accomplished by partitioning the information stream into segments, identifying occurrences of the substitution data pattern and the forbidden data pattern in a respective segment, constructing a flag for each occurrence, and replacing all occurrences of the forbidden data pattern in that segment with the substitution data pattern. Decoding keys, substitution data patterns, substitution flags or any other information needed to recover the original information is assembled with the modified information in a form that does not equal the forbidden data pattern.

Some embodiments provide a three-dimensional (3D) circuit structure that has two or more vertically stacked bonded layers with a machine-trained network on at least one bonded layer. As described above, each bonded layer can be an IC die or an IC wafer in some embodiments with different embodiments encompassing different combinations of wafers and dies for the different bonded layers. The machine-trained network in some embodiments includes several stages of machine-trained processing nodes with routing fabric that supplies the outputs of earlier stage nodes to drive the inputs of later stage nodes. In some embodiments, the machine-trained network is a neural network and the processing nodes are neurons of the neural network. In some embodiments, one or more parameters associated with each processing node (e.g., each neuron) is defined through machine-trained processes that define the values of these parameters in order to allow the machine-trained network (e.g., neural network) to perform particular operations (e.g., face recognition, voice recognition, etc.). For example, in some embodiments, the machine-trained parameters are weight values that are used to aggregate (e.g., to sum) several output values of several earlier stage processing nodes to produce an input value for a later stage processing node.

The invention provides a time domain integrated temperature sensor. Charging time of a capacitor is controlled through a shaped clock signal, so that the capacitor generates a charging time delay signal related with an input clock cycle. A time delay signal produced due to the fact that a path of PTAT currents charge the capacitor in the preceding control mode and a time delay signal produced due to the fact that a path of CTAT currents charge the capacitor in the same mode generate a pulse signal through an XOR operation, the width of the pulse signal is related with temperature and the input clock cycle, the pulse width of the pulse signal is quantized through the same input clock signal, and the obtained quantized result is completely counteracted with relevance of the input clock cycle, namely, an output value of the temperature sensor has no relation with the cycle of the input clock signal. The problem that due to the fact that an existing time domain integrated temperature sensor changes with the clock signal cycle, readings are inconsistent is solved, and precision of the time domain integrated temperature sensor is improved to a certain extent.

A non-volatile logic gate, including a magnetic material having a shape induced magnetic anisotropy, wherein a shape of the magnetic material has a first vertex, a second vertex, and a third vertex and supports a single magnetic domain; regions of the magnetic material including a first input region adjacent the first vertex, a second input region adjacent the second vertex, and an output region adjacent a third vertex; the first input region for receiving a first logic input to the logic gate, the second input region for receiving a second logic input to the logic gate, and the output region for outputting at least one logic output of the logic gate; and the shape induced magnetic anisotropy determining at least part of a truth table for the logic gate, so that the logic gate produces the at least one logic output from the logic inputs using the shape.

A filter network designed for providing high frequency selectivity with a high degree of reliability and availability. The filter network comprises a superconducting filter and a non-superconducting filter, or a combination thereof to form multiplexers. A receive side of the non-superconducting filter pre-filters received RF signals before inputting them to the superconducting filter. The non-superconducting filter is constructed and arranged to pass RF signals having a frequency within a first pass band to the superconducting filter. The superconducting device is constructed and arranged to exhibit a high-degree of frequency selectivity in further narrowing the received RF signals. Other aspects are directed to the arrangement, construction, and uses of the same structures to accomplish different but similar goals. In a multiplexed configuration, various combinations of transmit filters are used to enable the use of a common antenna with the receive side electronics, which may be located at the top of the antenna tower or in the base station.

Provided are a semiconductor memory device and a method of generating a chip enable signal thereof. The device includes a plurality of memory chips and an interface chip that are stacked. Each of the memory chips includes a control signal setting unit, which sets input signals applied to first and second input nodes as less significant 2-bit control signals of n-bit control signals, performs a logic AND operation on the less significant 2-bit control signals to generate AND operated signals, performs a logic XOR operation on each of the AND operated signals and each bit signal of more significant n−2-bit input signals applied to third to n-th input nodes to set the n−2-bit control signals, outputs the signal applied to the second input node through a first output node, inverts the signal applied to the first input node to output the inverted signal through a second output node, and outputs the more significant n−2-bit input signals through third through n-th output nodes, respectively. The first through n-th output nodes of one of two adjacent memory chips are respectively connected to the first through n-th input nodes of the other of the two adjacent memory chips.

Various circuit techniques for implementing ultra high speed circuits use current-controlled CMOS (C³MOS) logic fabricated in conventional CMOS process technology. An entire family of logic elements including inverter/buffers, level shifters, NAND, NOR, XOR gates, latches, flip-flops and the like are implemented using C³MOS techniques. Optimum balance between power consumption and speed for each circuit application is achieve by combining high speed C³MOS logic with low power conventional CMOS logic. The combined C³MOS/CMOS logic allows greater integration of circuits such as high speed transceivers used in fiber optic communication systems.

It is configured by plurality of NAND circuits connected in series through a plurality of inverters, and a plurality of NOR circuits connected in series through the plurality of inverters. Each of a plurality of source signal lines provided in a pixel portion is connected to one input terminal of a NAND circuit and a NOR circuit, and an output of an inspection is obtained from final lines of the NAND circuit and the NOR circuit connected in series. In this manner, an inspecting circuit which is capable of determining a defect simply and accurately by using a small-scale circuit, and a method thereof are provided.

Methods and systems are described for receiving, over a plurality of consecutive signaling intervals, a plurality of codewords, each codeword received as a plurality of symbols via wires of a multi-wire bus, the plurality of symbols received at a plurality of multi-input comparators (MICs), wherein each symbol is received by at least two MICs, generating, for each codeword, a corresponding linear combination of the received symbols, generating a plurality of composite skew measurement signals over the plurality of consecutive signaling intervals, each composite skew measurement signal based on samples of one or more linear combinations, and updating wire-specific skew values of the wires of the multi-wire bus, wherein one or more wire-specific skew values are updated according to composite skew measurement signals associated with linear combinations formed by at least two different MICs.

Integrated circuit devices include data inversion circuits therein that are configured to evaluate at least first and second ordered groups of input data in parallel with an ordered group of output data previously generated by the data inversion circuit. The data inversion circuit is further configured to generate inverted versions of the first and second ordered groups of input data as versions of the first and second ordered groups of data in parallel at outputs thereof whenever a number of bit differences between the first ordered group of input data and the ordered group of output data is greater than one-half a size of the first ordered group of input data and a number of bit differences between the second ordered group of input data and the version of the first ordered group of input data is greater than one-half a size of the second ordered group of input data, respectively.