Abstract: A capacitive interposer (caposer) is disposed inside an integrated circuit package between a die and an inside surface of the package. Conductive layers within the caposer constitute a bypass capacitor. In a through-hole caposer, micro-bumps on the die pass through through-holes in the caposer and contact corresponding landing pads on the package. As they pass through the caposer, power and ground micro-bumps make contact with the plates of the bypass capacitor. In a via caposer, power and ground micro-bumps on the die are coupled to power and ground landing pads on the package as well as to the plates of the bypass capacitor by power and ground vias that extend through the caposer. In signal redistribution caposer, conductors within the caposer redistribute signals between die micro-bumps and package landing pads. In an impedance matching caposer, termination structures within the caposer provide impedance matching to a printed circuit board trace.

Abstract: A capacitive interposer (caposer) is disposed inside an integrated circuit package between a die and an inside surface of the package. Conductive layers within the caposer constitute a bypass capacitor. In a through-hole caposer, micro-bumps on the die pass through through-holes in the caposer and contact corresponding landing pads on the package. As they pass through the caposer, power and ground micro-bumps make contact with the plates of the bypass capacitor. In a via caposer, power and ground micro-bumps on the die are coupled to power and ground landing pads on the package as well as to the plates of the bypass capacitor by power and ground vias that extend through the caposer. In signal redistribution caposer, conductors within the caposer redistribute signals between die micro-bumps and package landing pads. In an impedance matching caposer, termination structures within the caposer provide impedance matching to a printed circuit board trace.

Abstract: An FPGA is laid out as a plurality of repeatable tiles, wherein the tiles are disposed in columns that extend from one side of the die to another side of the die, and wherein each column includes tiles primarily of one type. Because substantially all die area of a column is due to tiles of a single type, the width of the tiles of each column can be optimized and is largely independent of the size of the other types of tiles on the die. The confines of each type of tile can therefore be set to match the size of the circuitry of the tile. Rather than providing a ring of input/output blocks (IOBs) around the die periphery, IOB tiles are disposed in columns only. Where more than two columns worth of IOBs is required, more than two columns of IOB tiles can be provided.

Abstract: A device for queuing information combines the speed of SRAM with the low cost and low power consumption of DRAM, affording substantial expansion of high-speed data storage in queues without corresponding increases in costs. The queues have a variable size, and provide fast, flexible and efficient data storage via an SRAM interface and a DRAM body. The queues may hold pointers to buffer addresses or other data that allow manipulation of information in the buffers via manipulation of the queues. Particular utility for this mechanism exists in situations for which high-speed access to queues is beneficial, flexible queue size is advantageous, and/or the smaller size and lower cost of DRAM compared to SRAM is of value.

Abstract: An intelligent network interface card (INIC) or communication processing device (CPD) works with a host computer for data communication. The device provides a fast-path that avoids protocol processing for most messages, greatly accelerating data transfer and offloading time-intensive processing tasks from the host CPU. The host retains a fallback processing capability for messages that do not fit fast-path criteria, with the device providing assistance such as validation even for slow-path messages, and messages being selected for either fast-path or slow-path processing. A context for a connection is defined that allows the device to move data, free of headers, directly to or from a destination or source in the host. The context can be passed back to the host for message processing by the host. The device contains specialized hardware circuits that are much faster at their specific tasks than a general purpose CPU.

Abstract: An intelligent network interface card (INIC) or communication processing device (CPD) works with a host computer for data communication. The device provides a fast-path that avoids protocol processing for most messages, greatly accelerating data transfer and offloading time-intensive processing tasks from the host CPU. The host retains a fallback processing capability for messages that do not fit fast-path criteria, with the device providing assistance such as validation even for slow-path messages, and messages being selected for either fast-path or slow-path processing. A context for a connection is defined that allows the device to move data, free of headers, directly to or from a destination or source in the host. The context can be passed back to the host for message processing by the host. The device contains specialized hardware circuits that are much faster at their specific tasks than a general purpose CPU.