Abstract: The invention provides a simple interface circuit between a large capacity, high speed DRAM and a single port SRAM cache to achieve fast-cycle memory performance. The interface circuit provides wider bandwidth internal communications than external data transfers. The interface circuit schedules parallel pipeline operations so that one set of data buses can be shared in cycles by several data flows to save chip area and alleviate data congestion. A flexible design is provided that can be used for a range of bandwidths of data transfer and generally any size bandwidth ranging from 32 to 4096 bits wide can use this same approach.

Abstract: A improved method of making of silicon on sapphire structure and/or device is disclosed. In a first preferred embodiment, a single silicon oxide layer is placed between the silicon layer and the sapphire layer. This can be done by attaching the silicon oxide layer on the silicon layer, e.g. by growing or depositing, and then attaching the sapphire layer to the oxide layer using wafer bonding. In an alternative embodiment, a first silicon oxide layer is attached to the silicon layer, e.g. by growing or depositing. A second silicon oxide layer is then attached to the sapphire layer, e.g. by depositing. Then the first and second silicon oxide layers are attached by a wafer bonding technique.

Abstract: A fuse latch array system for an embedded DRAM (eDRAM) having a micro-cell architecture, a wide data bandwidth and wide internal bus width is disclosed for localizing all the fuse information for redundancy replacement purposes. The fuse latch array system includes a fuse latch array having a plurality of memory cells where fuse information is scanned therein sequentially or parallel, or a combination thereof to be compatible with conventional fuse latch scanning protocols, during power-on. When the fuse information is stored in the fuse latch array, it is accessed as a page during a page mode operation. The accessed page contains column redundancy information corresponding to the active bank. The fuse latch array is decoded by row and column, so that the memory cell corresponding to the active bank can be easily located, even if there are thousands of banks within the eDRAM.

Abstract: An improved silicon on sapphire structure and/or device has one or more buffer layers. In a first preferred embodiment, the buffer layer is layer of silicon oxide material that prevents the stress induced defects in the silicon layer. In an alternative embodiment, the buffer layer comprises two layers. A first silicon oxide layer attached to the silicon to insure a perfect interface between the silicon. A second silicon oxide layer then is attached to the sapphire layer. The first and second silicon oxide layers are then attached, e.g., by a wafer bonding technique. This structure has no conductive paths beneath the oxide insulator(s) and therefore enables improved performance in radio frequency applications.

Abstract: Apparatus and method for providing high dielectric constant decoupling capacitors for semiconductor structures. The high dielectric constant decoupling capacitor can be fabricated by depositing high dielectric constant material between adjacent conductors on the same level, between conductors in successive levels, or both, to thereby provide very large capacitance value without any area or reliability penalty.

Abstract: A memory system having a plurality of T-RAM cells arranged in an array is presented where each T-RAM cell has dual vertical devices and is fabricated over a SiC substrate. Each T-RAM cell has a vertical thyristor and a vertical transfer gate. The top surface of each thyristor is coplanar with the top surface of each transfer gate within the T-RAM array to provide a planar cell structure for the T-RAM array. A method is also presented for fabricating the T-RAM array having the vertical thyristors, the vertical transfer gates and the planar cell structure over the SiC substrate.

Abstract: A fuse latch array system for an embedded DRAM (eDRAM) having a micro-cell architecture, a wide data bandwidth and wide internal bus width is disclosed for locatizing all the fuse information for redundancy replacement purposes. The fuse latch array system includes a fuse latch array having a plurality of memory cells where fuse information is scanned therein sequentially or parallel, or a combination thereof to be compatible with conventional fuse latch scanning protocols, during power-on. When the fuse information is stored in the fuse latch array, it is accessed as a page during a page mode operation. The accessed page contains column redundancy information corresponding to the active bank. The fuse latch array is decoded by row and column, so that the memory cell corresponding to the active bank can be easily located, even if there are thousands of banks within the eDRAM.

Abstract: A micromachined electromechanical random access memory (MEMRAM) array is disclosed which includes a plurality of MEM memory cells, where each MEM memory cell has an MEM switch and a capacitor. The MEM switch includes a contact portion configured for moving from a first position to a second position for reading out a charge stored within the capacitor or for writing the charge to the capacitor. A method is also disclosed for fabricating each MEM memory cell of the MEMRAM array.

Abstract: The speed of memories is increased by trading memory density (or area) for speed (or cycle time). An n by n memory array is used to reduce the memory cycle time by 1/n. For example, if an existing memory cycle time is 6 ns, in order to achieve a 3 ns (or n=2) cycle time, a 2 by 2 memory array is used. Or, in order to achieve a 1 ns cycle time (or n=6), then a 6 by 6 memory array is used.

Abstract: A method and structure for producing metallic polymer conductor lines comprising of an alternative methodology to a traditional damascene approach, called a cloisonne or inverse damascene approach. The cloisonne approach comprises the steps of coating a photosensitive polymer such as pyrrole or aniline with a silver salt on a semiconductor substrate. Using standard photolithography and resist developing techniques, the conducting polymer is exposed to a wet chemical developer, removing a portion of the exposed conducting polymer region, leaving only conducting polymer lines on top of the substrate. Next, an insulating dielectric layer is deposited over the entire structure and a chemical mechanical polish planarization of the insulator is performed creating the conducting polymer lines. Included in another aspect of the invention is a method and structure for a self-planarizing interconnect material comprising a conductive polymer thereby reducing the number of processing steps relative to the prior art.

Abstract: A semiconductor device is presented which includes a self-aligned, planarized thin-film transistor which can be used in various integrated circuit devices, such as static random access memory (SRAM) cells. The semiconductor device has a first field-effect transistor and a second field-effect transistor. The second field-effect transistor overlies the first field-effect transistor, and the first field-effect transistor and the second field-effect transistor share a common gate. The second field-effect transistor includes a source and a drain which are self-aligned to the shared gate in a layer of planarized seminconductor material above the first field-effect transistor. In one embodiment, the second field-effect transistor is a thin-film transistor, and the shared gate has a U-shape wrap-around configuration at a body of the thin-film transistor.

Abstract: A system and method for generating random noise for use in testing electronic devices comprises a first random pattern generator circuit for generating first sets of random bit pattern signals; one or more delay devices each receiving a trigger input signal and a random bit pattern signal set for generating in response a respective delay output signal, each delay output signal being delayed in time with respect to a respective trigger signal, a delay time being determined by the bit pattern set received; and, an oscillator circuit device associated with a respective one or more delay devices for receiving a respective delay output signal therefrom and generating a respective oscillating signal, each oscillator signal generated being used to generate artificial random noise for emulating a real noise environment in an electronic device.

Abstract: Thermal cooling structures of diamond or diamond-like materials are provided for conducting heat away from semiconductor devices. A first silicon-on-insulator embodiment comprises a plurality of thermal paths, formed after shallow trench and device fabrication steps are completed, which extend through the buried oxide and provide heat dissipation through to the underlying bulk silicon substrate. The thermal conduction path material is preferably diamond which has high thermal conductivity with low electrical conductivity. A second diamond trench cooling structure, formed after device fabrication has been completed, comprises diamond shallow trenches disposed between the devices and extending through the buried oxide layer. An alternative diamond thermal cooling structure includes a diamond insulation layer deposited over the semiconductor devices in either an SOI or bulk silicon structure.

Abstract: A data retention system has master-slave latches for holding data in an active mode; a data retention latch for preserving data read from the master latch in a sleep mode, which is connected to the master latch in parallel with the slave latch; a first multiplexer for receiving data externally provided and feedback data from the data retention latch, and selectively outputting either the data externally provided or the feedback data to the master latch in response to a first control signal; and a second multiplexer for transferring output data of the master latch to the slave latch and the data retention latch in response to a second control signal, wherein power for the data retention latch remains turned on in the sleep mode, while power for the data retention system except for the data retention latch is turned off.

Abstract: A power management circuit for an SRAM system including one or more isolated memory arrays and implementing a power source including a local power supply associated with each memory array and an external power supply connected to local supplies during an active mode of operation. The power management circuit comprises: a switch mechanism for disconnecting the external power supply to each of local power supply during a low power mode of operation; and, a refresh timing circuit implementing memory array refresh operation by selectively connecting the external power supply to a respective local power supply during the low power mode. During the low power mode, the refresh circuit intentionally enables the local power supply to float and allow it to drift to a lower predetermined voltage level prior to the memory array refresh operation.

Abstract: A static redundancy arrangement for a circuit using a focused ion beam anti-fuse methodology which reduces the circuit layout area and the switching activity compared to a prior art dynamic redundancy scheme, resulting in less power, a simpler design and higher speed. Focused ion beam anti-fuse methodology is used to program redundancy for circuits, particularly wide I/O embedded DRAM macros. An anti-fuse array circuit is comprised of a plurality of anti-fuse programming elements, each of which comprises a latch circuit controlled by a set input signal, and an anti-fuse device which is programmed by a focused ion beam.

Abstract: A power management circuit for an SRAM system including one or more isolated memory arrays and implementing a power source including a local power supply associated with each memory array and an external power supply connected to the local supplies during an active mode of operation. The power management circuit comprises: a switch mechanism for disconnecting the external power supply to each of local power supply during a low power mode of operation; and, a refresh timing circuit implementing memory array refresh operation by selectively connecting the external power supply to a respective local power supply during the low power mode. During the low power mode, the refresh circuit intentionally enables the local power supply to float and allow it to drift to a lower predetermined voltage level prior to the memory array refresh operation.

Abstract: A chip packaging system and method for providing enhanced thermal cooling including a first embodiment wherein a diamond thin film is used to replace at least the surface layer of the existing packaging material in order to form a highly heat conductive path to an associated heat sink. An alternative embodiment provides diamond thin film layers disposed on adjacent surfaces of the chip and the chip package. Yet another alternative embodiment includes diamond thin film layers on adjacent chip surfaces in a chip-to-chip packaging structure. A final illustrated embodiment provides for the use of an increased number of solder balls disposed in at least one diamond thin film layer on at least one of a chip and a chip package joined with standard C4 technology.

Abstract: A voltage control system and methodology for maintaining internally generated voltage levels in a semiconductor chip. The method comprises the steps of intermittently sampling an internal voltage supply level during a low power or “sleep” mode of operation; comparing the internal voltage supply level against a predetermined voltage reference level; and, activating a voltage supply generator for increasing the internal voltage supply level when the internal voltage supply level falls below the predetermined voltage reference level. The voltage supply generator is subsequently deactivated when the voltage supply level is restored to the predetermined voltage reference level. The sampling cycle may be appropriately tailored according to chip condition, chip temperature, and chip size. In one embodiment, the voltage control system and methodology is implemented in DRAM circuits during a refresh operation.

Abstract: A new micro-cell redundancy scheme for a wide bandwidth embedded DRAM having a SRAM cache interface. For each bank of micro-cell array units comprising the eDRAM, at least one micro-cell unit is prepared as the redundancy to replace a defected micro-cell within the bank. After array testing, any defective micro-cell inside the bank is replaced with a redundancy micro-cell for that bank. A fuse bank structure implementing a look-up table is established for recording each redundant micro-cell address and its corresponding repaired micro-cell address. In order to allow simultaneous multi-bank operation, the redundant micro-cells may only replace the defective micro-cells within the same bank. When reading data from eDRAM, or writing data to eDRAM, the micro-cell array address is checked against the look-up table to determine whether that data is to be read from or written to the original micro-cell, or the redundant micro-cell.