Abstract: A network switch includes an input layer to receive a data stream with a set of cells. Each cell includes data and a header to designate a destination device. The input layer includes a set of input layer circuits. A selected input layer circuit of the set of input layer circuits receives the data stream. The selected input layer circuit includes a set of queues corresponding to a set of destination devices. The selected input layer circuit is configured to assign a selected cell from the data stream to a selected queue of the set of queues. The selected queue corresponds to a selected destination device specified by the header of the selected cell. An intermediate layer includes a set of intermediate layer circuits, each intermediate layer circuit has a set of buffers corresponding to the set of destination devices.

Abstract: A content search system includes multiple pipelined search engines that implement different portions of a regular expression search operations. For some embodiments, the search pipeline includes a DFA engine, an NFA engine, and a token stitcher that combines partial match results generated by the DFA and NFA engines in a manner that prevents either engine from becoming a bottleneck. In addition, the token stitcher can be configured to implement unbounded sub-expressions without utilizing resources of the DFA or NFA engines.

Abstract: Embodiments of the invention relate to programmable data register circuits and programmable clock generation circuits For example, some embodiments include a buffer circuit for receiving input data and sending output data signals along a series of signal lines with a signal strength, and a signal modulator configured to determine the signal strength based on a control input. Some embodiments include a clock generation circuit for receiving clock reference and sending output clock signals along a series of signal lines with a signal character, and a signal modulator configured to determine the signal character based on a control input.

Abstract: A content addressable memory (CAM) (100) can include a CAM memory array (102) having both a data field (102-0) and a mask field (102-1). A multiplexer (MUX) (108) can selectively load data from either a register (104) or an external data input (106) to one or both fields (102-0 and 102-1) of CAM memory array (102).

Abstract: A processing system includes an interchip interface that comprises an interchip interface module having an arbiter to allocate a dedicated time slice in every fixed number of time slices, to assign a first priority to store data item(s) from a first-type channel having a first datapath width in memory during the dedicated time slice. In the remaining time slices of the fixed number of time slices, the arbiter further arbitrates among multiple channels of one or more types other than a first type, where the multiple channels correspond to at least one datapath width different from the first datapath width, and channels with wider datapath win the arbitration. The arbiter further arbitrates among two or more channels of the same type if a certain type of channel(s) wins the arbitration in a time slice. A method for implementing the same is also disclosed.

Abstract: Systems, methods, and circuits provide phase inversion of a clock signal. A first clock signal is received. A phase inversion signal indicates the existence of a 180 degree phase difference between the first clock signal and a second clock signal. As a result of the phase inversion signal indicating the 180 degree phase difference, the system, methods and circuits adapt the first clock signal by extending the first clock signal by a phase such that the first clock signal's rising edges and falling edges align with the second clock signal's rising edges and falling edges. As a result, the 180 degree phase difference between the clock signals is eliminated.

Abstract: A network-on-chip system, method, and computer program product are provided for transmitting messages utilizing a centralized on-chip shared memory switch. In operation, a message is sent from one of a plurality of agents connected on a messaging network. The message is received at a central shared memory switch, the central shared memory switch being in communication with each of the plurality of agents. Further, the message is transmitted from the central shared memory switch to a destination agent, the destination agent being one of the plurality of agents.

Abstract: A frequency synthesis circuit includes a phase locked loop and an interpolator circuit. The phase locked loop circuit receives a reference clock and a feedback clock and generates an output clock with a frequency based on the reference clock and the feedback clock. An interpolator circuit is coupled in the feedback path of the phase locked loop circuit. An interpolator control circuit generates an interpolator control word that specifies a variable time delay for the interpolator circuit. The interpolator circuit receives the output clock, and generates the feedback clock by introducing a variable time delay in the output clock in accordance with the interpolator control word. The time variable delay varies the frequency of the output circuit. Embodiments for frequency synthesis circuits that include a spread spectrum frequency clock generator, frequency modulators, and a fixed frequency clock generator circuit are disclosed.

Abstract: An adaptive distortion reduction system comprising: an input interface to receive a distorted signal comprising a distorted component and an undistorted component, the distorted component being at least in part attributed to an exogenous signal; and an adaptive distortion reduction module coupled to the input interface, to perform linearization based at least in part on the distorted signal and information associated with the exogenous signal, to obtain a corrected signal that is substantially similar to the undistorted component; wherein the adaptive self-linearization module includes: a first digital signal processor (DSP) that is adapted to obtain a filter transfer function that approximates a transfer function to be corrected; and a second DSP that is configured using configuration parameters of the first DSP.

Abstract: A receiver is optimized by adapting parameters of a linear equalizer component within the receiver. Data decisions and error decisions are generated. These data decision and error decisions are used to derive an error rate of data by measuring the number of margin hits that occur. A balance value is also calculated from the data decisions and the error decisions. The balance value is used to update parameters of the linear equalizer. The updating of the parameter continues until the number of margin hits has been minimized.

Abstract: A system, method, and computer program product are provided for conditionally sending a request for data to a node based on a determination. In operation, a first request for data is sent to a cache of a first node. Additionally, it is determined whether the first request can be satisfied within the first node, where the determining includes at least one of determining a type of the first request and determining a state of the data in the cache. Furthermore, a second request for the data is conditionally sent to a second node, based on the determination.

Abstract: A CAM-based search engine is disclosed that reduces power consumption during a plurality of different search operations concurrently performed in a plurality of device pipelines by selectively applying one of a number of different power reduction techniques for each pipeline in response to configuration data indicating the type of search operation that is being performed in the pipeline.

Abstract: A circuit for clocking includes an input data path, a receiver, a set of flip-flops, at least one interpolator and a controller. The receiver is coupled to the input data path for receiving input data. The flip-flops, coupled to the receiver, sample the input data. A first interpolator, coupled to one or more of the flip-flops, receives the sampled input data. The controller, coupled to the first interpolator, controls the first interpolator by providing phase information regarding the input data to the first interpolator. The circuit reduces any jitter transferred from the input path to an output path.

Abstract: A CAM cell is disclosed that can be selectively configured to store either base-2 data words or base-3 data words. When configured to store base-3 data words, the quaternary CAM cell compares 3 comparand bits representative of a base-3 comparand value with the base-3 data value stored in the CAM cell. Storing base-3 data words in such CAM cells increases the data storage density of associated CAM arrays.

Abstract: A search system for detecting whether one or more overlapping sequences of input characters match a regular expression including a prefix string preceding an intermediate expression having a quantified number m of characters belonging to a specified character class is disclosed. The search system includes a CAM array for storing the regular expression, a shift register for counting sequences of input characters that match the character class, and a control circuit that enables the shift register in response to a prefix match and increments the shift register in response to character class matches.

Abstract: Quaternary CAM cells are provided that include a compare circuit having a discharge path between a match line and ground potential, the single discharge path consisting essentially of a single transistor. In an embodiment, the single transistor has a gate coupled to a pull-down node and the compare circuit includes first and second gating transistors connected in series between the pull-down node and a ground potential, the first gating transistor having a gate to receive a comparand bit, and the second gating transistor having a gate to receive a complemented comparand bit.

Abstract: A digital signal processing system comprising: an input terminal to receive an input signal that includes a distorted component and an undistorted component, the input signal having a sampling rate of R; and an adaptive self-linearization module coupled to the input terminal, to perform self-linearization based at least in part on the input signal to obtain an output signal that is substantially undistorted, wherein: the adaptive self-linearization module is to generate a replica distortion signal that is substantially similar to the distorted component, the generation being based at least in part on a target component having a sampling rate of R/L, L being an integer greater than 1; the adaptive self-linearization module includes a first digital signal processor (DSP) that is adapted to obtain a filter transfer function that approximates a system distortion transfer function, and a second DSP that is configured using configuration parameters of the first DSP.

Abstract: A CAM device for comparing a search key with a plurality of ternary words stored in a CAM array includes one or more population counters, a pre-compare memory, and a pre-compare circuit. The present embodiments reduce the power consumption of CAM devices during compare operations between a search key and ternary words stored in a CAM array by selectively enabling the match lines in the CAM array in response to pre-compare operations between a set of population counts corresponding to the masked search key and a set of population counts corresponding to the ternary words stored in the CAM array.

Abstract: A frequency-domain based echo and NEXT canceller is provided. The canceller uses log2 encoding to precondition the error signal representing the echo. An improved gradient constraint is applied on at least a portion of a full weight vector in a least-mean-square algorithm. The least-mean-square algorithm is used to compute filter coefficients. The filter coefficients are multiplied by a frequency-domain data vector using a frequency-domain multiplier to generate frequency-domain output vector.

Abstract: An advanced processor comprises a plurality of multithreaded processor cores each having a data cache and instruction cache. A data switch interconnect is coupled to each of the processor cores and configured to pass information among the processor cores. A messaging network is coupled to each of the processor cores and a plurality of communication ports. The data switch interconnect is coupled to each of the processor cores by its respective data cache, and the messaging network is coupled to each of the processor cores by its respective message station. In one aspect of an embodiment of the invention, the messaging network connects to a high-bandwidth star-topology serial bus such as a PCI express (PCIe) interface capable of supporting multiple high-bandwidth PCIe lanes. Advantages of the invention include the ability to provide high bandwidth communications between computer systems and memory in an efficient and cost-effective manner.