Abstract: In accordance with the present invention a system for detecting transaction errors in a system comprising a plurality of data processing devices using a common system interconnect bus, comprises a node controller operably connected to said system interconnect bus and a plurality of interface agents communicatively coupled to said node controller. Error corresponding to transactions between said interface agents and other processing modules in said system are directed to said node controller; and wherein transaction errors that would not normally be communicated to said system interconnect bus are communicated by said node controller to said system interconnect bus to be available for detection. In an embodiment of the present invention, the interface agents operate in accordance with the hypertransport protocol. A system control and debug unit and a trace cache operably connected to the system bus can be used to diagnose and store errors conditions.

Abstract: A multiprocessor switching device substantially implemented on a single CMOS integrated circuit is described in connection with a descriptor-based packet processing mechanism for use in efficiently assigning and processing packets to a plurality of processors. A plurality of descriptors associated with each packet transfer are written back to memory in order, divided into subset groups and assigned to processors, where each processor searches the assigned subset for EOP and associated SOP descriptors to process.

Abstract: Electrically and thermally enhanced die-up ball grid array (BGA) packages are described. A BGA package includes a stiffener, substrate, a silicon die, and solder balls. The die is mounted to the top of the stiffener. The stiffener is mounted to the top of the substrate. A plurality of solder balls are attached to the bottom surface of the substrate. A top surface of the stiffener may be patterned. A second stiffener may be attached to the first stiffener. The substrate may include one, two, four, or other number of metal layers. Conductive vias through a dielectric layer of the substrate may couple the stiffener to solder balls. An opening may be formed through the substrate, exposing a portion of the stiffener. The stiffener may have a down-set portion. A heat slug may be attached to the exposed portion of the stiffener. A locking mechanism may be used to enhance attachment of the heat slug to the stiffener. The heat slug may be directly attached to the die through an opening in the stiffener.

Abstract: A multiprocessor switching device substantially implemented on a single CMOS integrated circuit is described in connection with a packet data transfer circuit that uses a fragment storage buffer to align and/or merge data being transferred to or from memory on a plurality of channels. In a packet reception embodiment, a data shifter and fragment store buffer are used to align received packet data to any required offset. The aligned data may and then be written to the system bus or combined with data fragments from prior data cycles before being written to the system bus. When packet data is being transferred to memory on a plurality of channels, the fragment storage may be channelized using register files or flip-flops to store intermediate values of packets and states for each channel.

Abstract: A multiprocessor switching device substantially implemented on a single CMOS integrated circuit is described in connection with a channelized timer for use in controlling the issuance of signals to the processor(s) or control logic (such as interrupts, descriptors, etc.) that that identify system-related functions for a plurality of channels. Using control registers to select an individual bit of a multi-bit counter, a timing interval pulse is provided for prompting signal generation that is otherwise subject to a minimum count requirement.

Abstract: A system and method for improving the bandwidth for data read and write operations in a multi-node system by using uncacheable read and write commands to a home node in the multi-node system so that the home node can determine whether the commands needs to enter the coherent memory space. In one embodiment where nodes are connected via HT interfaces, posted commands are used to transmit uncacheable write commands over the HT fabric to a remote home node so that no response is required from the home node. When both cacheable and uncacheable memory operations are mixed in a multi-node system, a producer-consumer software model may be used to require that the data and flag must be co-located in the home node's memory and that the producer write both the data and flag using regular HT I/O commands.

Abstract: A multiprocessor switching device substantially implemented on a single CMOS integrated circuit is described in connection with a parallel routing scheme for calculating routing information for incoming packets. Using the programmable hash and route routing scheme, a hash and route circuit can be programmed for a variety of applications, such as routing, flow-splitting or load balancing.

Abstract: An integrated receiver with channel selection and image rejection substantially implemented on a single CMOS integrated circuit is described. A receiver front end provides programmable attenuation and a programmable gain low noise amplifier. Frequency conversion circuitry advantageously uses LC filters integrated onto the substrate in conjunction with image reject mixers to provide sufficient image frequency rejection. Filter tuning and inductor Q compensation over temperature are performed on chip. The filters utilize multi track spiral inductors. The filters are tuned using local oscillators to tune a substitute filter, and frequency scaling during filter component values to those of the filter being tuned. In conjunction with filtering, frequency planning provides additional image rejection. The advantageous choice of local oscillator signal generation methods on chip is by PLL out of band local oscillation and by direct synthesis for in band local oscillator.

Abstract: An electrically and mechanically enhanced die-down tape substrate ball grid array (BGA) package substrate is described. An IC package includes a substrate that has a first surface. The first surface has a central opening. A stiffener/heat spreader has a first surface. The first surface of the stiffener has a central ground ring. The first surface of the stiffener is coupled to a second surface of the substrate. The central opening has an edge. The edge includes at least one of the following: (a) a protruding edge portion that extends across at least a portion of the central ground ring, (b) a recessed edge portion that exposes a portion of the central ground ring, or (c) a hole proximate to the edge, wherein the hole exposes a portion of the central ground ring.

Abstract: A voltage regulator includes a first stage capable of receiving a reference voltage and capable of having a first current flowing through the first stage. A second stage is capable of having a second current flowing through the second stage. A third stage is capable of outputting an output voltage and capable of having a third current flowing through the second stage. The first, second and third currents are proportional to each other throughout a range of operation of the voltage regulator between substantially zero output current and maximum output current. The first stage drives the second stage as a low input impedance load.

Abstract: A method of constructing a lookup table for network switch includes the steps of snooping a communication channel in the network switch for lookup table information. Upon detection of lookup table information on the communication channel, the lookup table information transmitting the lookup table information to a remote system memory. This constructs a lookup table in the remote system memory.

Abstract: Multi-carrier modulation fiber optic systems constructed using a series of electrical carriers, modulating the data on the electrical carriers and combining the carriers to form a wideband signal. The wideband signal can then be intensity modulated on a laser and coupled to a fiber optic channel. A receiver may then receive the laser signal from the fiber optic channel and convert it into an electrical signal. Multi-carrier modulation may be applied to existing fiber channels, which may be of lower quality. Existing fiber channels may have characteristics which prevent or restrict the transmission of data using intensity modulation at certain frequencies. An adaptive multi-carrier modulation transmitter may characterize an existing fiber optic channel and ascertain the overall characteristics of the channel. The transmitter and receiver can then be configured to use various bandwidths and various modulations in order to match the transfer characteristic of the fiber channel.

Abstract: The present invention is directed to a seal structure and a method for forming a seal structure for a semiconductor die. An elongate region which is electrically isolated from the remainder of the substrate, such as a well region of a conductivity type opposite that of the substrate, extends around the major portion of the periphery of the substrate. A gap is left between the two ends of the elongate region along the minor portion of the periphery of the substrate not covered by the elongate region. A conductive seal ring is formed around the periphery of the substrate at the elongate region and spans the gap between the ends of the elongate region. The substrate of the semiconductor die is only brought into electrical contact with the seal ring at the gap between the ends of the elongate region.

Abstract: A graphics integrated circuit chip is used in a set-top box for controlling a television display. The graphics chip processes analog video input, digital video input, and graphics input. The chip includes a single polyphase filter that preferably provides both anti-flutter filtering and scaling of graphics. Anti-flutter filtering may help reduce display flicker due to the interlaced nature of television displays. The scaling of graphics may be used to convert the normally square pixel aspect ratio of graphics to the normally rectangular pixel aspect ratio of video.

Abstract: Various systems and methods providing signal delay compensation for circuits such as a multi-pair gigabit Ethernet transceiver are disclosed. In an analog implementation a buffer with an adjustable delay may be used to minimize the delay mismatch between clock trees. The delay of the adjustable-delay buffer is controlled by bias voltages that determine the charging and discharging current to the adjustable buffer. A phase detector circuit is used to compare the clock phases for rising and falling edges, and to adjust the bias voltages that control these edges. In a digital implementation a selector switch, responsive to a phase detector, may be used to route clock signals through circuit elements to delay clock signals.

Abstract: A method of filtering data packets in a network device is disclosed. An incoming packet is received from a port and the incoming packet is inspected and packet fields are extracted. The incoming packet is classified based on the extracted packet fields and action instructions are generated. The incoming packet is then modified based on the action instructions. Further, the inspection and extraction includes applying inspection mask windows to any portion of the incoming packet to extract programmable packet fields.

Abstract: An apparatus for providing a programmable gain attenuator (PGA) while minimizing the influence of semiconductor switches on the signal being attenuated. An example apparatus comprises a impedance ladder with taps forming the junctions between impedances the PGA is then programmed by grounding the taps through terminating resistors.

Abstract: A method for distributing sets of collision resolution parameters to be used for resolution of network access contention events among nodes of a non-centralized media access control shared medium network. A set of collision resolution parameters is provided which includes a sequence of fixed numbers for resolving a single network access contention event. A single collision signal slot master node is identified when one or more candidate collision signal slot master nodes exist. Collision signal slot request messages are sent from client nodes addressed to all network nodes. Collision signal slot assignment messages are sent from the master node to the client nodes. A collision resolution parameter set to be employed by that given client node is obtained at a given client node from within a received collision signal slot assignment message. Collision signal slot acknowledgment messages are sent from client nodes addressed to all network nodes.

Abstract: In one embodiment, a computer readable medium comprises at least first instructions and second instructions. The first instructions, when executed, compute a first plurality of routes. Each route of the first plurality of routes corresponds to a respective net of a plurality of nets in an integrated circuit layout, and represents a theoretically optimal route of the respective net according to a graph theory based algorithm. The second instructions, when executed, compare each of the first plurality of routes to a corresponding route of a current plurality of routes, each of the current plurality of routes corresponding to the respective net of the plurality of nets and currently existing in the integrated circuit layout. A method is also contemplated.

Abstract: An apparatus and process for relabelling and redirecting at least some of the read transaction data frames and the write transaction write data and transfer ready frames in a network so as to bypass a storage manager and pass directly between the client and a storage device via a switch. This eliminates the storage manager as a bottleneck. Some embodiments redirect every read and write transaction, and others redirect only large transactions, or only ones not stored in cache or when latency gets too high. Redirection is accomplished by transmission from the storage manager to a switch redirection commands that contain old and new address data. When a frame to be redirected comes in, its address is compared to the old address data. If there is a match, the new address data is substituted and the frame is passed to a conventional routing process to be routed so as to bypass the storage manager.