Abstract: An embodiment of an embedded cache memory in an image sensor comprises a memory cell array wherein the memory cells are substantially isolated from laterally adjacent memory. The memory cell array includes a plurality of memory cells. Each of the memory cells is formed in a standard CMOS image sensor process without the need for SOI processes. Each cell includes first and second n-type and p-type regions arranged around a vertically integrated gate. Data is written to a cell by causing carriers to accumulate in the body of the device through carrier generation mechanisms that may include impact ionization, band-to-band tunneling and/or channel-initiated secondary hot electrons.

Abstract: An integrated circuit is provided comprising an array of memory cells connected by word and bit lines, respectively, wherein each memory cell comprises a thyristor structure, an anode terminal that connects the thyristor structure with a respective bit line, a gate terminal that connects the thyristor structure with a respective word line, and a cathode terminal. The integrated circuit further comprises a drive/sensing circuitry configured to apply a first sequence of voltage signals at the anode terminal and the gate terminal, wherein the voltage signals are defined with respect to the cathode terminal. The first sequence comprises a first voltage signal at the anode terminal, a second voltage signal at the gate terminal, and thereafter a combination of a third voltage signal at the anode terminal and a fourth voltage signal at the gate terminal, wherein the third voltage signal is lower than the first voltage signal and lower than the fourth voltage signal.

Abstract: A semiconductor device includes a memory cell including a thyristor element with a gate having a pnpn structure formed in a semiconductor substrate, and a plurality of access transistors formed on the semiconductor substrate and each connected at a first terminal thereof to a storage node at one terminal of the thyristor element such that a potential at the storage node can be transmitted to bit lines different from each other, the gate of the thyristor element and the gates of the plurality of access transistors of the memory cell being connected to word lines different from one another.

Abstract: A semiconductor device, includes a memory cell including a thyristor element with a gate having a pnpn structure formed on a semiconductor substrate and having first and second terminals, and an access transistor formed on the semiconductor substrate and having first and second terminals connected to a bit line and the first terminal of the thyristor element, respectively, and a control section including a load current element whose load current flows, upon reading out operation, to the second terminal side of the thyristor element and configured to carry out access control to the memory cell.

Abstract: A nonvolatile storage device includes a plurality of bit lines 21 arranged in a column direction on a substrate; a plurality of word lines 35 arranged in a row direction on the substrate; a memory cell array 20 having a plurality of memory cells 31, where a store state of each of the memory cells 31 changes according to an electric signal relatively applied to the word line 35 and the bit line 21; a word line selection unit having a needle 51 relatively movable with respect to the substrate which comes into contact with one word line 35, setting the word line 35 in contact with the needle 51 to a selection state; and a sense amplifier 48 detecting through the bit line an electrical signal exhibiting the store state of the memory cell 31 to be connected to the word line.

Abstract: A new memory cell can contain only a single thyristor. There is no need to include an access transistor in the cell. In one embodiment, the thyristor is a thin capacitively coupled thyristor. The new memory cell can be connected to word, bit, and control lines in several ways to form different memory arrays. Timing and voltage levels of word, bit and control lines are disclosed.

Abstract: A memory cell is provided which includes an access transistor and a gated lateral thyristor (GLT) device. The access transistor includes a source node. The gated lateral thyristor (GLT) device includes an anode node coupled to the source node of the access transistor.

Abstract: The present invention pertains to a circuit arrangement that, in one example, facilitates reading or determining an amount of current that flows through a memory cell when one or more voltages are applied to the cell. The amount of current resulting from the applied voltages is a function of the amount of charge stored within the cell, among other things, and the amount of stored charge represents information stored within the cell. As such, reading the resulting current allows data stored within the cell to be accessed and retrieved. It will be appreciated however, that use of the circuitry disclosed herein is not limited to memory applications. Rather, it can be used in any application where current sensing is required along with a regulated supply voltage.

Abstract: A present method of fabricating a memory device includes the steps of providing a dielectric layer;, providing an opening in the dielectric layer, providing a first conductive body in the opening, providing a switching body in the opening, the first conductive body and switching body filling the opening, and providing a second conductive body over the switching body. In an alternate embodiment, a second dielectric layer is provided over the first-mentioned dielectric layer, and the switching body is provided in an opening in the second dielectric layer.

Abstract: A memory device includes a plurality of word lines extending as rows and bit lines extending as columns. A memory cell is coupled between a word line and a bit line, wherein the memory cell includes a unipolar memory element selectively coupled to the bit line via a trigger element.

Abstract: One of the simplest forms of data storage devices is the diode array storage device. However, a problem with diode array storage devices is that as the size of the array increases, the number of non-addressed diodes connected between a given selected row or column of the array and the non-addressed columns or rows of the array, respectively, also becomes very large. While the leakage current through any one non-addressed diode on the selected row or column will have little impact on the operation of the device, the cumulative leakage through multiple thousands of non-addressed diodes can become significant. This aggregate leakage current can become great enough that the output voltage can be shifted such that the threshold for distinguishing between a one state and a zero state of the addressed diode location can become obscured and can result in a misreading of the addressed diode location. The present invention is a means to manage the leakage currents in a diode array storage device.

Abstract: A memory device using a nanotube cell comprises a plurality of nanotube sub-cell arrays each having a hierarchical bit line structure including a main bit line and a sub-bit line. In the memory device, a nanotube cell array comprising a capacitor and a PNPN nanotube switch which does not require an additional gate control signal is located between a word line and the sub-bit line, so that a cross point cell array is embodied to reduce the whole chip size.

Abstract: Switching operations, such as those used in memory devices, are enhanced using a semiconductor device having a thyristor adapted to switch between conducting and blocking states and operate at low power. According to an example embodiment of the present invention, thyristor characteristics are managed over a broad temperature range using a control circuit for coupling a signal, such as a DC voltage signal, to a portion of a thyristor for controlling temperature-related operation thereof, e.g., for controlling bipolar gains. In one implementation, a control port adaptively adjusts a signal coupled to the thyristor as a function of temperature, such that at relatively low temperatures unwanted increases in holding current (IH) are prevented. In another implementation, the control port couples the signal at relatively high temperature operation for controlling the forward blocking voltage (VFB) in such a manner that a blocking state of the thyristor is held.

Abstract: Non-Volatile Memory (NVM) cells include a selection circuit for providing an output based on selecting between an input data signal and an output of a Multiple Time Programmable (MTP) NVM element. The input data signal may be latched by a latch circuit such as a flip-flop first. The selector circuit's output is used to confirm the programming values for the MTP NVM element such that the element can be programmed correctly without losing time by reading the programmed MTP NVM element or reprogramming a misprogrammed element.

Abstract: A nonvolatile storage device includes a plurality of bit lines 21 arranged in a column direction on a substrate; a plurality of word lines 35 arranged in a row direction on the substrate; a memory cell array 20 having a plurality of memory cells 31, where a store state of each of the memory cells 31 changes according to an electric signal relatively applied to the word line 35 and the bit line 21; a word line selection unit having a needle 51 relatively movable with respect to the substrate which comes into contact with one word line 35, setting the word line 35 in contact with the needle 51 to a selection state; and a sense amplifier 48 detecting through the bit line an electrical signal exhibiting the store state of the memory cell 31 to be connected to the word line.

Abstract: A multiple-bit memory cell for use in a magnetic random access memory circuit includes a first adiabatic switching storage element having a first anisotropy axis associated therewith and a second adiabatic switching storage element having a second anisotropy axis associated therewith. The first and second anisotropy axes are oriented at a substantially non-zero angle relative to at least one bit line and at least one word line corresponding to the memory cell. The memory cell is configured such that two quadrants of a write plane not used for writing one of the storage elements can be beneficially utilized to write the other storage element so that there is essentially no loss of write margin in the memory cell.

Abstract: A re-writable memory with multiple memory layers. Using both terminals of a memory cell in a stacked cross point structure for selection purposes allows multiple layers of conductive lines to be selected as long as there is only one memory cell that has two terminals selected. Sharing logic over multiple layers allows driver sets to be reused.

Abstract: A thyristor-based memory may comprise a thyristor accessible via an access transistor. A temperature dependent bias may be applied to at least one of a supporting substrate and an electrode capacitively-coupled to a base region of the thyristor. The voltage level of the adaptive bias may change with respect to temperature and may influence and/or compensate an inherent bipolar gain of the thyristor in accordance with the change in bias and may enhance its performance and/or reliability over a range of operating temperature. In a particular embodiment, the thyristor may be formed in a layer of silicon of an SOI substrate and the adaptive bias coupled to a supporting substrate of the SOI structure.

Abstract: A first and a second set of memory cells are connected to the same first word line and second word line. At the commencement of data writing, the first word line is set up. The first set of memory cells is read and temporarily stored into a buffer. At about the same time, the bit lines of the second set of memory cells is set up. After completion of reading of the first set of memory cells, the bit lines of this set of memory cells are set up (while the setting up of the bit lines of the second set of memory cells continues). After the bit lines of both sets of memory cells are set up, the second word line is pulsed. At this time, written into both sets of memory cells begins, which comprises data previously read from the first set of memory cells and new data to be written into the second set of memory cells. It is found that this method reduces the overall write time.

Abstract: Device and method for memory cell isolation. The memory cell includes a resistive component, such as a magnetic random access memory (MRAM) cell, and an isolation component, such as a four-layer diode. The memory cell may be included in a memory array. The method includes rapidly applying a forward bias across the isolation element to activate the isolation element.

Abstract: A re-recordable data storage medium is disclosed. The medium in one embodiment includes a phase-changeable layer and an intermediate layer. A junction between the intermediate layer and another layer of the medium provides a conduction barrier under no illumination that is substantially diminished under illumination of the regions of the phase-changeable layer that are in the appropriate phase.

Abstract: A re-writable memory with multiple memory layers. Using both terminals of a memory cell in a stacked cross point structure for selection purposes allows multiple layers of conductive lines to be selected as long as there is only one memory cell that has two terminals selected. Sharing logic over multiple layers allows driver sets to be reused.

Abstract: A magnetoresistive film has at least a first magnetic layer, a second magnetic layer, a nonmagnetic layer, a third magnetic layer, and a fourth magnetic layer stacked in the order named. In the magnetoresistive film, at least the first magnetic layer contains Gd and the fourth magnetic layer contains Tb and/or Dy, each of the first magnetic layer and fourth magnetic layer has an easy axis of magnetization along a perpendicular direction to a film plane, and the second magnetic layer and the third magnetic layer have a greater spin polarization than the first magnetic layer and the fourth magnetic layer. Furthermore, the second and third magnetic layers are magnetic layers containing at least Co and Co contents thereof are not less than 20 at. % nor more than 90 at. %.

Abstract: A chalcogenide material is formed to a first thickness over the first conductive electrode material. The chalcogenide material includes AxBy. A layer that includes a metal is formed to a second thickness over the chalcogenide material. The metal including layer defines some metal including layer transition thickness for the first thickness of the chalcogenide material such that when said transition thickness is met or exceeded, said metal including layer when diffused within said chalcogenide material transforms said chalcogenide material from an amorphous state to a crystalline state. The second thickness being less than but not within 10% of said transition thickness. The metal including layer is irradiated effective to break a chalcogenide bond of the chalcogenide material and diffuse at least some of the metal into the chalcogenide material.

Abstract: A content addressable memory (CAM) cell includes a memory cell storing data values. The memory cell includes a surrounding-gate thyristor and an access transistor. The CAM cell also includes a compare circuit coupled among the memory cell and a match line. The compare circuit receives data and comparand data and affects a logical state of a match line in response to a predetermined relationship between the data and comparand data. The compare circuit includes a first transistor set coupled for conduction state control by signals representative of the data, and a second transistor set coupled for conduction state control by signals representative of the comparand data.

Abstract: A nonvolatile semiconductor memory device includes a memory array in which a plurality of memory cells are arranged in a row direction and a column direction, each of the memory cells being formed by connecting one end of a variable resistive element for storing information according to a change in electric resistance caused by an electric stress and a drain of a selection transistor to each other on a semiconductor substrate, a voltage switch circuit for switching among a program voltage, an erase voltage and a read voltage to be applied to the source line and the bit line connected to the memory cell, and a pulse voltage applying circuit.

Abstract: Switching times of a thyristor-based semiconductor device are improved by enhancing carrier drainage from a buried thyristor-emitter region. According to an example embodiment of the present invention, a conductive contact extends to a doped well region buried in a substrate and is adapted to drain carriers therefrom. The device includes a thyristor body having at least one doped emitter region buried in the doped well region. A conductive thyristor control port is adapted to capacitively couple to the thyristor body and to control current flow therein. With this approach, the thyristor can be rapidly switched between resistance states, which has been found to be particularly useful in high-speed data latching implementations including but not limited to memory cell applications.

Abstract: A novel capacitively coupled NDR device can be used to implement a variety of semiconductor circuits, including high-density SRAM cells and power thyristor structures. In one example embodiment, the NDR device is used as a thin vertical PNPN structure with capacitively-coupled gate-assisted turn-off and turn-on mechanisms. An SRAM based on this new device is comparable in cell area, standby current, architecture, speed, and fabrication process to a DRAM of the same capacity. In one embodiment, an NDR-based SRAM cell consists of only two elements, has an 8 F2 footprint, can operate at high speeds and low voltages, has a good noise-margin, and is compatible in fabrication process with main-stream CMOS. This cell significantly reduces standby power consumption compared to other types of NDR-based SRAMs.

Abstract: A semiconductor device having a thyristor is arranged in a manner that reduces or eliminates manufacturing difficulties commonly experienced in the formation of such devices, as well as facilitates the implementation of the semiconductor device in a variety of applications. According to an example embodiment of the present invention, a thyristor is formed having some or all of the body of the thyristor extending above a substrate surface of a semiconductor device. The semiconductor device includes at least one transistor having source/drain regions formed in the substrate prior to the formation of the thyristor. One or more layers of material are deposited on the substrate surface and used to form a portion of a body of the thyristor that includes anode and cathode end portions. Each end portion is formed having a base region and an emitter region, and at least one of the end portions includes a portion that is in the substrate and electrically coupled to the transistor.

Abstract: A control signal MBPRG is inputted to individual block decoders that constitute a block decoder section 37 of an ACT type flash memory. Then, the level of the control signal MBPRG is set to “H” to select all the blocks regardless of the contents of address signals a5 through a13, and one word line WL is selected from all the blocks by the addresses a0 through a4. By thus selecting one word line WL every block that is electrically separated by the select transistor and simultaneously applying a write voltage during the test to the same number of word lines WL as the number of blocks, the possible occurrence of a bad influence exerted on the other memory cells is prevented even when the memory cells in which a write operation during the test has been executed include a memory cell that has a negative threshold voltage.

Abstract: Area efficient static memory cells and arrays containing p-n-p-n transistors which can be latched in a bistable on state. Each transistor memory cell includes a gate which is pulse biased during the write operation to latch the cell. Also provided is a CMOS fabrication process to create the cells and arrays.

Abstract: A novel capacitively coupled NDR device can be used to implement a variety of semiconductor circuits, including high-density SRAM cells and power thyristor structures. In one example embodiment, the NDR device is used as a thin vertical PNPN structure with capacitively-coupled gate-assisted turn-off and turn-on mechanisms. An SRAM based on this new device is comparable in cell area, standby current, architecture, speed, and fabrication process to a DRAM of the same capacity. In one embodiment, an NDR-based SRAM cell consists of only two elements, has an 8F2 footprint, can operate at high speeds and low voltages, has a good noise-margin, and is compatible in fabrication process with main-stream CMOS. This cell significantly reduces standby power consumption compared to other type of NDR-based SRAMs.

Abstract: A novel capacitively coupled NDR device can be used to implement a variety of semiconductor circuits, including high-density SRAM cells and power thyristor structures. In one example embodiment, the NDR device is used as a thin vertical PNPN structure with capacitively-coupled gate-assisted turn-off and turn-on mechanisms. An SRAM based on this new device is comparable in cell area, standby current, architecture, speed, and fabrication process to a DRAM of the same capacity. In one embodiment, an NDR-based SRAM cell consists of only two elements, has an 8 F2 footprint, can operate at high speeds and low voltages, has a good noise-margin, and is compatible in fabrication process with main-stream CMOS. This cell significantly reduces standby power consumption compared to other types of NDR-based SRAMs.

Abstract: Area efficient static memory cells and arrays containing p-n-p-n or n-p-n-p transistors which can be latched-up in a bistable on state. Each transistor memory cell includes a gate which is pulse biased during the write operation to latch-up the cell. Also provided are linked memory cells in which the transistors share common regions.

Abstract: A manufacturing method and a structure for ROM component having a silicon controlled rectifier as the basic memory instead of a channel transistor in a conventional ROM, and using a formation of contact windows for coding a ROM instead of performing an ion implantation process. Also, since a silicon controlled rectifier occupies a smaller component surface area, the level of integration is correspondingly increased. Furthermore, due to interposition of an insulating layer between two bit lines, short circuiting between the adjacent bit lines is prevented. The component of this invention operates by applying a suitable voltage to the word line electrode and the bit line electrode respectively to select a particular memory unit, and as a result, a current will flow in a vertical direction through the memory unit, exit through the common electrode depending on the ON/OFF state of the memory, and be detected there.

Abstract: The SRAM cell configuration has at least six transistors in each memory cell. Four of the transistors form a flip-flop and they are arranged at the corners of a quadrilateral. The flip-flop is driven by two of the transistors, which are disposed so as to adjoin diagonally opposite corners of the quadrilateral and outside the quadrilateral. Adjacent memory cells along a word line can be arranged in such a way that a first bit line and a second bit line of the adjacent memory cells coincide. The transistors are preferably vertical and are arranged at semiconductor structures (St1, St2, St3, St4, St5, St6) produced from a layer sequence. Two of the transistors having n-doped channel regions are preferably formed in each case on two semiconductor structures.

Abstract: A magnetic random access memory (10) has a plurality of stacked memory cells on semiconductor substrate (11), each cell basically having a portion of magnetic material (12), a word line (13), and sense line (14). Upper sense line (22) is electrically coupled to lower sense line (12) via conductor line (23) with ohmic contacts. In order to read and store states in the memory cell, lower and upper word lines (13, 18) are activated, thereby total magnetic field is applied to portion of magnetic material (11). This stacked memory structure allows magnetic random access memory (10) to integrate more memory cells on semiconductor substrate (11).

Abstract: A semiconductor memory has bit lines, word lines, ground lines, and memory cells. The bit lines intersect the word and ground lines, to form intersections where the memory cells are arranged, respectively. Each of the memory cells consists of a double-emitter transistor. This transistor has a collector, a first emitter, and a second emitter. Each base-emitter junction of the transistor has an N-shaped negative differential current-voltage characteristic that shows a relatively small gain up to a peak current and a relatively large gain after a valley current. The first emitter of each transistor is connected to a corresponding one of the ground lines. The second emitter is connected to a corresponding one of the word lines. The collector is connected to a corresponding one of the bit lines. Each of the memory cells has a small number of elements and requires only a small area.

Abstract: A semiconductor static random access memory cell is implemented by four field effect transistors formed on a silicon layer over a buried silicon oxide layer and two resistors formed in an inter-level insulating structure; additional capacitors are formed under the buried silicon oxide layer, and are respectively connected to the gate electrodes of the field effect transistors serving as driving transistors of the memory cell so as to enhance the stability of the memory cell without increase the transistor size.

Abstract: A memory cell with vertically stacked crossovers. In prior memory cells, crossover connections within the memory cell were implemented in the same device layer. This wasted valuable design space, since the crossovers were therefore required to sit side-by-side in the layout design. The present invention implements crossovers in different materials on different device layers. The crossovers may therefore be vertically stacked on top of each other, reducing the area of the memory cell.

Abstract: A random access memory (RAM) cell in 6H-SiC having storage times when all bias is removed long enough to be considered nonvolatile. The nonvolatile random access memory (NVRAM) cell comprises a bit line, a charge storage device in silicon carbide, and a transistor in silicon carbide connecting the charge storage device to the bit line. The bipolar NVRAM cell has a bipolar transistor with a base region, an emitter region, and a floating collector region, wherein the charge storage device in the bipolar NVRAM is a p-n junction adjacent the floating collector region of the cell. The metal-oxide-semiconductor (MOS) NVRAM has a MOS field effect transistor (MOSFET) with a channel region, a source region, and a drain region, wherein the charge storage device in the MOS NVRAM is a MOS capacitor adjacent the drain region of the MOSFET.

Abstract: A new static memory cell based on the bistable operation of a three-terminal four layer semiconductor device working in the forward blocking state is disclosed. The power consumption of the memory cell is low. The switching speed of the memory cell is in the nanosecond range. The memory cell may be integrated into VLSI processes and is of a size suitable for VLSI applications. The memory cell comprises a semiconductor device which comprises: an n-type semiconductor cathode region; a p-type semiconductor gate region; a third semiconductor region adjacent the gate region; a fourth region adjacent the third semiconductor region; and a hole-injecting boundary between the third semiconductor region and the fourth region. The semiconductor device is preferably a p-n-p-n device. The gate current-voltage characteristics of the device comprise a negative resistance region. The device is designed to operate in a forward blocking low current state.

Abstract: A CMOS SRAM cell includes "true" and "false" NMOS word-line access transistors, "true" and "false" NMOS pull-down transistors, and "true" and "false" PMOS pull-down transistors arranged in a classical six-transistor SRAM electrical configuration. "True" and "false" inter-level interconnects of silicidable material provide for respective five-way connections among the transistors. The "true" inter-level interconnect connects: the drain of the "true" pull-up transistor, a gate level polysilicon conductor defining and connecting the gates of the "false" pull-up transistor and the "false" pull-down transistor, and a diffusion region defining and connecting the source of the "true" access transistor and the drain of the "true" pull-down transistor.

Abstract: A vertically integrated DRAM cell having a storage time of at least 4.5 hours at room temperature, formed from a wide-bandgap semiconductor such as GaAs or AlGaAs, in which an n-p-n bipolar access transistor is merged with a p-n-p storage capacitor, with the middle p-n layers being common to both. Similarly, a p-n-p transistor can be merged with an n-p-n storage capacitor.

Abstract: Bipolar transistor memory cell and method for use in a random access memory. A pair of state elements are cross coupled so that they assume opposite states in accordance with signals applied thereto, a pair of bipolar pass transistors are connected to respective ones of the state elements for applying signals to the state elements, and current flow through the pass transistors is monitored to determine the states of the state elements.

Abstract: This disclosure relates to a programmable write-once, read-only semiconductor memory array which has an improved current source for each bit line and an improved current sink for each Word line. This programmable write-once, read-only semiconductor memory array utilizes a SCR (PNPN or NPNP) on the end of each Word line of the array to function as a current sink to minimize voltage drop on the Word line and a SCR (PNPN or NPNP) on each Bit line of the array for current sourcing purposes. This disclosure also relates to an integrated SCR (PNPN or NPNP) for use with a plurality of connected semiconductor devices to provide either a current sourcing or current sinking or drawing function for the plurality of connected semiconductor devices.

Abstract: An electrical data storage element which provides for alteration of this stored data by way of a data line and address lines. Data is held as a charge set on a charge storage device. The state of the switch elements is sensed by way of the data line and address line.

Abstract: A process for controlling an optical pnpn thyristor to be driven comprises a step of applying a train of pulses to maintain a low impedance state of an optical pnpn thyristor which is shifted beforehand to be in the low impedance state by a positive set pulse. Each of the train of pulses is less than in its level than the positive set pulse. When light emission is required for the reading of an information, a positive set pulse is applied to the optical pnpn thyristor. As a result, electric power consumption is reduced during a time storing the information.

Abstract: In a semiconductor memory cell having PNPN type memory cells, a vertical PNPN element is used as a load transistor and a sense transistor or a hold transistor, or both. A buried layer is used as a wiring layer for a word line or a bit line, so that the switching speed can be increased and the memory cell area can be decreased.

Abstract: Bipolar memory arrays having lower quiescent leakage and higher switching speed are constructed by using coupled SCRs formed from vertical PNP and NPN devices. Buried collectors for the PNP and NPN devices are provided within the same isolation tub. A P type plug is used to connect the P collector of the PNP to the P base of the NPN in a region where the P base and P collector overlap. A single N epi-region serves as the base of the PNP and the collector of the NPN. The P plug is located within this N epi-region but part of the N epi-region adjacent to or around the P plug is left so that internal connection of the PNP base and NPN collector is not cut off by the P plug. The structure is particularly suited for use in large memory arrays. The method of fabrication is also described.