Abstract: A digital switching system comprises: (a) a line card layer containing a plurality of real or virtual line cards; (b) a switch card layer containing a plurality of real or virtual switch cards; and (c) an interface layer interposed between the line card layer and the switch card layer for providing serialization support services so that one or more of the line cards and switch cards can be operatively and conveniently disposed in a first shelf or on a first backplane that is spaced apart from a second shelf or from a second backplane supporting others of the line cards and/or switch cards. Such an arrangement allows for scalable expansion of the switching system in terms of number of lines served and/or transmission rates served. The flexibility of the system is owed in part to payload data being carried within payload-carrying regions of so-called ZCell signals as the payload data moves between the line card layer and the switch fabric layer.

Abstract: Fully-buffered dual in-line memory modules (FB-DIMM) include advanced memory buffers (AMBs) having enhanced skew, slew rate and output impedance control. The AMB includes user accessible registers that can be programmed to carefully control the edge placement (or phase) of signals generated from the AMB to multiple DRAMs on the module. This control of edge placement, which may be performed independently for each group of signals: clock (CLK, CLK#), command (RAS, CAS, WE), address (including bank address), data (DQ) and data strobe (DQS), provides 360 degrees of control (or one period). This means that any group of signals can be moved independently by one complete period relatively to any other group.

Abstract: The present invention relates to a method for forming a thin film resistor and a thin film resistor formed over a semiconductor substrate. A gate structure is formed and a dielectric layer is formed over the gate structure. A via is then etched that extends through the dielectric layer so as to expose a portion of the gate structure. A layer of titanium nitride is deposited using a chemical vapor deposition process. A rapid thermal anneal is performed in an oxygen ambient. The rapid thermal anneal incorporates oxygen into the titanium nitride, forming titanium oxynitride film. A layer of dielectric material is then deposited and etched-back to form a dielectric plug that fills the remaining portion of the via. The titanium oxynitride film is patterned to form a titanium oxynitride structure that is electrically coupled to the gate structure. A metal layer is deposited and patterned to form an interconnect structure that electrically couples the titanium oxynitride structure to other circuitry.

Abstract: An ATM switch according to the present invention includes a memory and a control circuit. The ATM switch generates a connection table in a memory, generates a multicast master entry including a limit field and a count field. The multicast master entry also includes address locations at which multicast ATM cells are stored. The ATM switch further generates one or more multicast member entries associated with said multicast master entry in said connection table, each multicast member entry identifying a destination connection on which said multicast ATM cells are to be transmitted. Further, the count field is initialized and the limit field is set at a predetermined value. The master entry is then determined to be active or inactive depending on a comparison between the count field and the limit field.

Abstract: An integrated circuit memory device includes a quad-port cache memory device and a higher capacity supplemental memory device. These memory devices operate collectively as a high speed FIFO having fast fall through capability and extended data capacity. The FIFO does not require complex arbitration circuitry to oversee reading and writing operations. The supplemental memory device may be an embedded on-chip memory device or a separate off-chip memory device (e.g., DRAM, SRAM). The quad-port cache memory device utilizes a data rotation technique to support bus matching. Error detection and correction (EDC) circuits are also provided to check and correct FIFO read data. The EDC circuits operate without adding latency to FIFO read operations.

Abstract: A voltage/current reference circuit includes a first bipolar transistor and a second bipolar transistor that exhibit a first voltage drop VBE1 and a second voltage drop VBE2, respectively. A first resistor, having a resistance R1, is configured to draw a first current equal to (VBE1?VBE2)/R1. A second resistor, having a resistance R2, is configured to draw a second current equal to VBE1/R2. A first transistor supplies the first and second currents to the first and second resistors. A second transistor, having a current mirror configuration with respect to the first transistor, directly provides a reference current equal to (VBE1?VBE2)/R1+VBE1/R2. A third transistor, having a current mirror configuration with respect to the first transistor, provides a current equal to the reference current to a third resistor having a resistance R3 and a third bipolar transistor that exhibits a third voltage drop VBE3, thereby generating a reference voltage.

Abstract: Content addressable memory (CAM) devices use both hard and soft priority techniques to allocate entries of different priority therein. The allocation of entries may change in response to additions or deletions of entries or as entries are reprioritized. The CAM devices include priority resolution circuits that can resolve competing soft and hard priorities between multiple hit signals that are generated in response to a search operation. Such hit signals may be active to reflect the presence of at least one matching entry within a CAM array block. The resolution of which active hit signal has the highest overall priority among many can be used to facilitate the identification of the location (e.g., array address and row address) of a highest priority matching entry within the entire CAM device. A priority resolution circuit may also resolve competing hard priorities between two or more active hit signals having equivalent soft priority.

Abstract: Gate stacks with sidewall spacers having improved profiles to suppress or eliminate void formation between the gate stacks during gap-filling is disclosed, along with a method of forming the gate structures over a semiconductor substrate. A gate dielectric layer is formed on a semiconductor substrate. Then, a gate stack 24 having a sidewall is formed over the gate dielectric layer. The gate stack 24 comprises a conductive layer 28 and a hard mask 30 overlying the conductive layer 28. A liner 32 is selectively deposited over the gate stack 24 such that the liner 32 is deposited on the hard mask 30 at a rate lower than the rate of deposition on the conductive layer 28. Thus, the liner 32 is substantially thinner on the hard mask 30 than on the conductive layer 28. A nitride spacer is formed over 34 the liner 32. A PMD layer is formed over the resultant structure, filling the gaps between adjacent gate stacks and substantially free of voids.

Abstract: A synchronous memory circuit is capable of double data transfer rate per clock cycle, 100% bus utilization (i.e., no idle clock cycles in bus turn arounds), and has only one clock cycle of latency in each of read and write burst operations.

Abstract: Clock processing logic and method for determining clock signal characteristics in reference voltage and temperature varying environments are described. A sample vector is characterized by bit locations corresponding to sequentially increasing delay values so that values stored in such bit locations indicate clock signal edges where value transitions occur. In one embodiment, edge detection logic and sensitivity adjustment logic are used in determining the clock period from such a sample vector. In another embodiment, an edge filter, sample accumulation logic, and clock period and jitter processing logic are used in determining an average clock period and clock jitter from a predefined number of such sample vectors.

Abstract: A network switch port includes a cell memory, a queuing system, a data path controller and an output buffer. The data path controller stores incoming cells derived from network data transmissions in the cell memory. The queuing system generates the cell memory address of each stored cell when the cell is to be forwarded from the cell memory, and the data path controller appends the cell memory address of that cell to a linked list of addresses of cells to be forwarded from the memory. When the linked list is not empty, the data path controller forwards cells from the cell memory to the output buffer in the order that their cell memory addresses were appended to the linked list. The output buffer stores and then sequentially forwards the cells outward from the switch port to a receiving network component which store them in a cell buffer until it can forward them elsewhere.

Abstract: Content addressable memory (CAM) devices include CAM logic that is configured to pass an instruction received at an instruction input port to an instruction output port without inspection or alteration. This enables the CAM devices to be operated as equivalent devices within a cascaded chain of CAM devices that collectively form multiple databases within a lookup engine having distributed CAM control. This CAM logic may include an input instruction register that is configured to latch the instruction received at the instruction input port and an output instruction register that is configured to latch the instruction received from the input instruction register. This CAM logic may also include an instruction FIFO that is configured to buffer instructions received from the input instruction register.

Abstract: An input or output switch port for a network switch converts each incoming packet into a cell sequence stores each cell in a cell memory. The switch port includes a traffic manager for queuing cells for departure from the cell memory and then signaling the cell memory to read out and forward cells in the order they are queued. The traffic manager selectively queues cells for departure on either a cell-by-cell or sequence-by-sequence basis. When cells are queued for departure on a cell-by-cell basis, cells of two or more sequences may be alternately read out and forwarded from the cell memory. Thus cells of different sequences may be interleaved with one another as they depart the cell memory. When a cell sequence is queued on a sequence-by-sequence basis all of its cells are read out of the cell memory and forwarded as a contiguous sequence and are not interleaved with cells of other sequences of the same departure queue.

Abstract: Impedance-matched output driver circuits utilize predriver circuits with analog control to provide enhanced operating characteristics. This analog control may be provided by an analog loop containing differential amplifiers that set the resolution limit of the output driver circuit. These output driver circuits include a first PMOS pull-up transistor having source and drain terminals electrically connected in series in a pull-up path of the output driver circuit. An NMOS pass transistor has a first current carrying terminal electrically connected to a gate terminal of the first PMOS pull-up transistor and a second current carrying terminal configured to receive a P-type analog reference voltage (VP). This P-type reference voltage controls the conductivity of the first PMOS pull-up transistor in the pull-up path. A gate terminal of the NMOS pass transistor is responsive to a pull-up data input signal (DINP).

Abstract: Content addressable memory (CAM) devices include at least one CAM array having a plurality of columns of ternary CAM cells therein. A bit/data line driver circuit, which receives search words and write words, is electrically coupled to the CAM array. The bit/data line driver circuit is configured to save power by logically combining incoming search words with a global mask that designates locations of active ones of the plurality of columns having exclusively locally masked CAM cells at valid entries therein. The bit/data line driver circuit includes a global mask generator configured to receive write words to be added to the CAM array during respective write operations.

Abstract: A delay-locked loop (DLL) circuit includes a phase interpolator circuit and variable delay circuit coupled in cascade and operative to generate an output clock signal that is delayed with respect to a reference clock signal responsive to respective first and second control signals applied to the phase interpolator and the variable delay circuit. The DLL circuit further includes a phase control circuit that generates the first and second control signals responsive to the output clock signal and the reference clock signal. The variable delay circuit may provide a coarser resolution than the phase interpolator circuit, for example, the variable delay circuit may include a tapped delay chain circuit configured to provide step changes in delay responsive to the second control signal.

Abstract: A content addressable memory (CAM) system that includes a row of NAND-type CAM cells divided into a plurality of segments. Each segment includes a plurality of series-connected switching transistors, wherein each of the switching transistors is part of a corresponding NAND-type CAM cell. The series-connected switching transistors of each segment are coupled to the series-connected switching transistors in an adjacent segment by a repeater circuit, thereby forming a chain of series-connected switching transistors and repeater circuits. A match line driver circuit is coupled to one end of the chain, and a match line is connected to the other end of the chain. If a match condition exists for the entire row, then a signal driven by the match line driver is propagated to the match line, through the chain of series-connected switching transistors and repeater circuits.

Abstract: A multi-port memory cell includes a first SRAM element having a first pair of access transistors electrically coupled to a pair of FIFO write bit lines. A second dual-port SRAM element is also provided. This second dual-port SRAM element has a second pair of access transistors electrically coupled to a pair of FIFO read bit lines and a third pair of access transistors electrically coupled to a pair of memory read bit lines. A direct path data transfer circuit is provided. This transfer circuit is configured to support a unidirectional data transfer path that extends from first storage nodes within the first SRAM element to second storage nodes within the second dual-port SRAM element. This transfer circuit is also responsive to a direct path word line signal.

Abstract: A hardware hashing circuit is configured to perform a hashing function on a received character string, thereby creating a hashed output value and a collision resolution value. A content addressable memory (CAM) receives the hashed output value, and in response, provides an index value and activates a hit signal if the hashed output value matches an entry of the CAM. A random access memory (RAM) receives the index value from the CAM. The RAM stores a collision resolution value and information associated with the character string in an entry associated with the index value. The RAM provides this information and collision resolution value in response to the index value. Logic circuitry indicates a collision if the hit signal is activated and the collision resolution value provided by the hardware hashing circuit does not match the collision resolution value provided by the RAM.

Abstract: A CMOS structure and a process for forming CMOS devices are disclosed in which gate film stacks are formed over a semiconductor substrate. A barrier layer and a first dielectric film are formed such that they extend over the gate film stacks. Metal lines are formed over the pre-metal dielectric film and spacers are formed that extend on opposite sides of the metal lines. A second dielectric film is formed that extends over the metal lines. A masking structure is formed that defines a contact opening. Selective etch processes are performed to form a self-aligned contact opening, with the adjacent metal lines and spacers aligning the self-aligned contact opening between adjacent gate film stacks. A metal layer is then deposited and planarized to form a self-aligned contact. The masking structure can also define additional contact openings, which are simultaneously etched and filled with metal to form borderless, strapped and shared contacts.