Abstract: A technique implements a novel high-speed high-density packet buffer utilizing a combination of high-speed and low-speed memory devices. The novel packet buffer is organized as a plurality of FIFO queues where each FIFO queue is associated with a particular input or output line. Each queue comprises a high-speed cache portion that resides in high-speed memory and a low-speed high-density portion that resides in low-speed high-density memory. The high-speed cache portion contains FIFO data that contains head and/or tail associated with the novel FIFO queue. The low-speed high-density portion contains FIFO data that is not contained in the high-speed cache portion.

Abstract: A technique synchronizes data retrieved from memory devices at a memory controller of a high-speed memory subsystem. Each memory device is organized into a plurality of data groupings. The memory controller stores (via one or more write operations) a known synchronization (sync) pattern at each data grouping on the memory devices and then retrieves (via one or more read operations) that sync pattern from the groupings. Synchronization logic located at a local clock boundary of the memory controller is configured to recognize the retrieved sync pattern and “automatically” synchronize all pieces of data retrieved from the data groupings, even though there may be substantial skew between the groupings.

Abstract: A technique for implementing a novel high-speed high-density packet buffer utilizing a combination of high-speed and low-speed memory devices. The novel packet buffer is organized as a plurality of first-in-first-out (FIFO) queues where each FIFO queue is associated with a particular input or output line. Each queue comprises a high-speed cache portion that resides in high-speed memory and a low-speed high-density portion that resides in low-speed high-density memory. Each high-speed cache portion contains FIFO data that contains head and/or tail information associated with a corresponding FIFO queue. The low-speed high-density portion contains FIFO data that is not contained in the high-speed cache portion. A queue identifier (QID) directory refills the high-speed portion of one or more queues with data from a corresponding low-speed portion. Queue head start and end offsets are used to determine whether a corresponding queue is empty.

Abstract: The present invention incorporates a technique that enables the processing of memory requests without requiring memory devices that support a request identifier (ID). The present invention maintains an association between a request identifier and a first memory request issued by a requestor and directed to a memory location. The first memory request contains an address corresponding to the memory location. A memory device corresponding to the address is selected. A second memory request without the identifier is issued to the selected memory device to access information stored at the location. The information is received from the memory device and associated with the request identifier.

Abstract: A coherency resolution technique enables efficient resolution of data coherency for packet data associated with a service queue of an intermediate network node. The packet data is enqueued on a write buffer prior to being stored on an external packet memory of a packet memory system. The packet data may be interspersed among other packets of data from different service queues, wherein the packets are of differing sizes. In response to a read request for the packet data, a coherency operation is performed by coherency resolution logic on the data in the write buffer to determine if any of its enqueued data can be used to service the request.

Abstract: In a processing system having a main memory, wherein information is stored in a compressed format for the purpose of gaining additional storage through compression efficiencies, a method and apparatus for providing compressed data integrity verification to insure detection of nearly any data corruption resulting from an anomaly anywhere in the logical processing or storage of compressed information. A cyclic redundancy code (CRC) is computed over a compressed data block as the data enters the compressor hardware, and the CRC is appended to the compressor output block before it is stored into the main memory. Subsequent read access results in comparing the CRC against a recomputation of the CRC as the block is uncompressed from the main memory. Any CRC miscompare implies an uncorrectable data error condition that may be used to interrupt the system operation.

Abstract: A highly-efficient system and methodology for organizing, storing and/or transmitting compressed data that achieves optimum compression throughput, enhances overall data compressibility, and reduces decompression latency.

Abstract: A highly-efficient system and methodology for organizing, storing and/or transmitting compressed data that achieves optimum compression throughput, enhances overall data compressibility, and reduces decompression latency.

Abstract: A method of maintaining cache coherency in a computer system including two or more processors sharing information, the processors coupled by two or more interconnects to a memory such that the processors are not directly coupled to the same is disclosed interconnect is disclosed.