Abstract: A semiconductor device including: a first memory cell including a non-volatile first variable resistance element that stores data by varying a resistance value and a selection transistor that selects the first variable resistance element; a first memory layer provided with more than one such first memory cell arranged in a plane; a second memory cell including a non-volatile second variable resistance element that stores data by varying a resistance value and a selection diode that selects the second variable resistance element; and a second memory layer provided with more than one such second memory cell arranged in a plane; wherein more than one such second memory layer is stacked over the first memory layer.

Abstract: A method for synchronizing two connection nodes by a reception node of the connection nodes with a clock data recovery circuit that generates a synchronization clock from input data. The method includes performing a synchronization process to establish synchronization between the connection nodes based on the synchronization clock, performing a connection failure process when the synchronization is not established when a first time elapses after receiving the input data, correcting the clock data recovery circuit when the synchronization is not established when a second time elapses after receiving the input data, wherein the second time is shorter than the first time, and performing a resynchronization process to establish synchronization between the connection nodes based on a synchronization clock, which is generated by the clock data recovery circuit that has been corrected, before the first time elapses and after the second time elapses.

Abstract: A memory is disclosed that can operate in a normal mode of operation or a testing mode of operation. In the testing mode of operation, the memory can measure various benchmarks of performance, such as read speed. The memory can perform an asynchronous read operation to read a word of electronic data that corresponds to an address or a page read operation in which multiple asynchronous read operations are performed to read multiple words of electronic data, also referred to as a page of electronic data, that correspond to multiple addresses. The memory can measure a time required, referred to as read speed, to read the word of electronic data or the multiple words of electronic data from the memory.

Abstract: Error correcting systems, methods, and devices for non-volatile memory are disclosed. In one embodiment, a non-volatile memory device comprises a data area for storing data, an error correcting code generation section for generating an error correcting code in response to receipt of a code generation command, and an error correcting code area for storing the error correcting code.

Abstract: A semiconductor device includes a semiconductor chip, a connection electrode including a first land electrode electrically coupled with the semiconductor chip, and a through electrode formed on an upper surface of the first land electrode to be electrically coupled with the first land electrode using a stud bump, and a sealing resin, through which the connection electrode passes, for sealing the semiconductor chip.

Abstract: Embodiments of the present invention include an acoustic processing device and a method for traversing a Hidden Markov Model (HMM). The acoustic processing device can include a senone scoring unit (SSU), a memory device, a HMM module, and an interface module. The SSU is configured to receive feature vectors from an external computing device and to calculate senones. The memory device is configured to store the senone scores and HMM information, where the HMM information includes HMM IDs and HMM state scores. The HMM module is configured to traverse the HMM based on the senone scores and the HMM information. Further, the interface module is configured to transfer one or more HMM scoring requests from the external computing device to the HMM module and to transfer the HMM state scores to the external computing device.

Abstract: Embodiments of the present invention include a substrate package, a method for multi chip packaging, and a multi-chip package. For example, the substrate package includes a first set of reference markers and a second set of reference markers. The first set of reference markers is disposed on the substrate package, where the first set of reference markers is configured to provide a first alignment for positioning a first integrated circuit (IC) and a second alignment for positioning a second IC on the substrate package. Further, the second set of reference markers is disposed at a different location on the substrate package than the first set of reference markers, where the second set of reference markers is configured to provide confirmation of the first alignment and the second alignment.

Abstract: Embodiments of the present invention include an apparatus, method, and system for integrating data processing logic with memory. An embodiment of a memory integrated circuit is designed to execute a task on the data in a memory array within a memory integrated circuit. The memory integrated circuit can include a memory array, a data access component, a data holding component, and a logic component. The data access component can be coupled to the memory array and configured to provide an address to the memory array. The data holding component can be coupled to the memory array and configured to temporarily store the data in the memory array located at the address. The logic component can be coupled to both the data access component and the data holding component, and be configured to execute a task using data received from the data holding component. The logic component can include combinational or sequential logic.

Abstract: Embodiments of the present invention include a data storage device and a method for storing data in a hash table. The data storage device can include a first memory device, a second memory device, and a processing device. The first memory device is configured to store one or more data elements. The second memory device is configured to store one or more status bits at one or more respective table indices. In addition, each of the table indices is mapped to a corresponding table index in the first memory device. The processing device is configured to calculate one or more hash values based on the one or more data elements.

Abstract: Embodiments of the present invention include an acoustic processing device, a method for acoustic signal processing, and a speech recognition system. The speech processing device can include a processing unit, a histogram pruning unit, and a pre-pruning unit. The processing unit is configured to calculate one or more Hidden Markov Model (HMM) pruning thresholds. The histogram pruning unit is configured to prune one or more HMM states to generate one or more active HMM states. The pruning is based on the one or more pruning thresholds. The pre-pruning unit is configured to prune the one or more active HMM states based on an adjustable pre-pruning threshold. Further, the adjustable pre-pruning threshold is based on the one or more pruning thresholds.

Abstract: A memory device is provided including a substrate. A first dielectric layer is formed over the substrate. An isolation trench is formed in a portion of the substrate and the first dielectric layer. At least two charge storage elements are formed over the first dielectric layer on opposite sides of the isolation trench. A second dielectric layer is formed over the at least two charge storage elements. A control gate layer is formed over the second dielectric layer, where the isolation trench has a width suitable for reducing cross-coupling noise of charge storage elements, and where the at least two charge storage elements have a height suitable for providing sufficient gate coupling between the at least two charge storage elements and the control gate layer.

Abstract: According to one aspect of the embodiment, a linear regulator circuit includes an output transistor outputting an output current based on a input voltage, an error amplifier outputting a control signal based on an electric potential difference between an output voltage based on the output current and a reference voltage, a buffer circuit coupled between the error amplifier and the output transistor, and a drive capability adjustment circuit adjusting a load drive capability of the buffer circuit in synchronization with the output current.

Abstract: A semiconductor device with efficient heat dissipating structures is disclosed. The semiconductor device includes a first semiconductor chip that is flip-chip mounted on a first substrate, a heat absorption portion that is formed between the first semiconductor chip and the first substrate, an outer connection portion that connects the first semiconductor chip to an external device and a heat conduction portion formed between the heat absorption portion and the outer connection portion to dissipate heat generated by the first semiconductor chip.

Abstract: A source-sensing configuration for non-volatile memory devices to simultaneously read 2 bits in two different memory cells sharing a same word line is disclosed. In a first cell arrangement, a drain of a first read cell is biased and its source and that of two adjacent cells in a direction towards the second read cell are connected through source bit lines to a source sense amplifier. In a second cell arrangement, the drain of the second read cell is biased and its source and that of its two adjacent cells in a direction towards the first read cell are connected through source bit lines to a source sense amplifier. A memory cell acts as a cell pipe and joins together the first and second cell arrangements. Driving all six source bit lines simultaneously allows the 2 bits to be simultaneously read while maintaining currents due to pipe effect substantially minimized.

Abstract: Embodiments described herein generally relate to landing gate pads for contacts and manufacturing methods therefor. A bridge is formed between two features to allow a contact to be disposed, at least partially, on the bridge. Landing the contact on the bridge avoids additional manufacturing steps to create a target for a contact.

Abstract: Semiconductor devices and methods of manufacturing such devices are described herein. According to embodiments, the semiconductor device can be made by forming an dielectric layer at a first region and at a second region of a semiconductor substrate. A gate conductor layer is disposed over the dielectric formed in the first and the second regions of the semiconductor substrate, and the second region is masked. A split gate memory cell is formed in the first region of the semiconductor substrate with a first gate length. The first region is then masked, and the second region is etched to define a logic gate that has a second gate length. The first and second gate lengths can be different.

Abstract: Semiconductor devices and methods of manufacturing such devices are described herein. According to embodiments, a semiconductor device includes a memory gate disposed in a first region of the semiconductor device. The memory gate may include a first gate conductor layer disposed over a charge trapping dielectric. A select gate may be disposed in the first region of the semiconductor device adjacent to a sidewall of the memory gate. A sidewall dielectric may be disposed between the sidewall of the memory gate and the select gate. Additionally, the device may include a logic gate disposed in a second region of the semiconductor device that comprises the first gate conductor layer.

Abstract: Embodiments described herein generally relate to methods of manufacturing charge-trapping memory by patterning the high voltage gates before other gates are formed. One advantage of such an approach is that a thin poly layer may be used to form memory and low voltage gates while protecting high voltage gates from implant penetration. One approach to accomplishing this is to dispose the layer of poly, and then dispose a mask and a thick resist to pattern the high voltage gates. In this manner, the high voltage gates are formed before either the low voltage gates or the memory cells.

Abstract: A semiconductor device and method of making such device is presented herein. The method includes disposing a gate layer over a dielectric layer on a substrate and further disposing a cap layer over the gate layer. A first transistor gate is defined having an initial thickness substantially equal to a combined thickness of the cap layer and the gate layer. A first doped region is formed in the substrate adjacent to the first transistor gate. The cap layer is subsequently removed and a second transistor gate is defined having a thickness substantially equal to the thickness of the gate layer. Afterwards, a second doped region is formed in the substrate adjacent to the second transistor gate. The first doped region extends deeper in the substrate than the second doped region, and a final thickness of the first transistor gate is substantially equal to the thickness of the second transistor gate.

Abstract: A method of manufacturing a semiconductor device includes forming a silicon nitride film having an opening portion on a semiconductor substrate, forming a silicon oxide film on the silicon nitride film and on a side face of the opening portion, performing an etching treatment to the silicon oxide film so that a sidewall is formed on the side face of the opening portion, forming a trench on the semiconductor substrate with use of the sidewall and the silicon nitride film as a mask, and forming an insulating layer in the trench. The step of forming the silicon oxide film includes oxidizing the silicon nitride film with a plasma oxidation method or a radical oxidation method.