Abstract: A data processing system containing a monolithic network of cells with sufficient redundancy provided through direct logical replacement of defective cells by spare cells to allow a large monolithic array of cells without uncorrectable defects to be organized, where the cells have a variety of useful properties. The data processing system according to the present invention overcomes the chip-size limit and off-chip connection bottlenecks of chip-based architectures, the von Neumann bottleneck of uniprocessor architectures, the memory and I/O bottlenecks of parallel processing architectures, and the input bandwidth bottleneck of high-resolution displays, and supports integration of up to an entire massively parallel data processing system into a single monolithic entity.

Abstract: A data processing system containing a monolithic network of cells with sufficient redundancy provided through direct logical replacement of defective cells by spare cells to allow a large monolithic array of cells without uncorrectable defects to be organized, where the cells have a variety of useful properties. The data processing system according to the present invention overcomes the chip-size limit and off-chip connection bottlenecks of chip-based architectures, the von Neumann bottleneck of uniprocessor architectures, the memory and I/O bottlenecks of parallel processing architectures, and the input bandwidth bottleneck of high-resolution displays, and supports integration of up to an entire massively parallel data processing system into a single monolithic entity.

Abstract: A communications bus enables communication of data signals in a parallel processing system having a plurality of substantially identical cells, each cell having an access point for transmitting data signals into the communications bus. The communications bus includes a plurality of parallel channels, and at least one channel crossover point associated with each cell. Each crossover point implements a regular change in a channel order of the communications bus, such that each access point is coupled to a channel of the communications bus. Propagation delays are reduced by inserting buffers at regular intervals along the length of each channel. An output buffer at a downstream boundary of each power domain of the system prevents undesired currents due to voltage mismatch. The propagation direction of data signals away from the access point, and propagation of data to an adjacent downstream cell can be controlled to reduce bus traffic and power consumption.

Abstract: A massively parallel data processing system consisting of an array of closely spaced cells where each cell has direct output means as well as means for processing, memory and input. The data processing system according to the present invention overcomes the von Neumann bottleneck of uniprocessor architectures, the I/O and memory bottlenecks that plague parallel processors, and the input bandwidth bottleneck of high-resolution displays.

Abstract: A method, apparatus and computer-readable storage medium for regulating packet flow through a device such as a router with a switch fabric. Congestion information, such as statistics on bandwidth utilization, is collected for each of a plurality of queues at an egress stage of the device. Based on the bandwidth utilization statistics, computations are performed to evaluate a “discard probability” for each queue. This information is transmitted to the ingress stage, either periodically or at other controlled time periods, such as when the discard probability changes significantly. The ingress stage can then proceed with controllable transmission or non-transmission of packets to the switch fabric, depending on the queue for which the packet is destined and also depending on the discard probability for that queue. In this way, congestion can be avoided even before it even has a chance to occur.

Abstract: A method and system for regulating packet flow to a downstream entity capable of forwarding packets to a plurality of intermediate destinations. The method includes maintaining a database of queues, each queue in the database being associated with packets intended to be forwarded to a corresponding one of a plurality of final destinations via a corresponding one of the intermediate destinations. Each queue in the database is further associated with a state that is either active or inactive. Upon receipt of a message from the downstream entity indicating a reduced (increased) ability of a particular one of the intermediate destinations to accept packets intended to be forwarded to a particular one of the final destinations, the method provides for rendering inactive (active) the state of the queue associated with packets intended to be forwarded to the particular final destination via the particular intermediate destination.

Abstract: Method, apparatus and software for processing sets of routing information in a router having a plurality of memory units accessible via separate access paths. The sets of routing information are typically routes received from neighbour nodes. The method includes creating a plurality of non-identical routing information subsets from each received set of routing information, accessing the memory units via the separate access paths and storing the routing information subsets created from a common set of routing information in respective ones of the plurality of memory units, By providing a distributed memory architecture for storing routing information, an increase in a router's memory requirements can be met by increasing the number of memory units.

Abstract: A switch fabric implemented on a chip includes an array of cells and an I/O interface in communication with the array of cells permitting exchange of data packets between the array of cells and components external to the array of cells. Each cell communicates with at least one other cell of the array, thereby permitting exchange of data packets between the cells of the array. Each cell includes a memory for holding a plurality of data packets for transmission to other cells of said array. Each data packet of the plurality of data packets has a characteristic element represented by a parameter, the parameter allowing one data packet to be distinguished from another data packet in the plurality of data packets. Each cell further includes a control entity operative to select at least one data packet from the plurality of data packets at least in part on a basis of the parameter and to transmit the selected data packet to another cell of said array of cells.

Abstract: A switch fabric implemented on a chip includes an array of cells and an I/O interface in communication with the array of cells for permitting exchange of data packets between the array of cells and components external to the array of cells. Each cell includes a transmitter in communication with the I/O interface and in communication with every other cell of the array, the transmitter being operative to process a data packet received from the I/O interface to determine a destination of the data packet and forward the data packet to at least one cell of the array selected on a basis of the determined destination. Each cell further includes a plurality of receivers associated with respective cells from the array, each receiver being in communication with a respective cell allowing the respective cell to forward data packets to the receiver, where the receivers are in communication with the I/O interface for releasing data packets to the I/O interface.

Abstract: A router includes a routing layer and a switching layer. The routing layer includes a plurality of I/O ports for exchanging data with components external to the router. The switching layer is adapted to switch data packets between I/O ports of the routing layer. The switching layer includes an array of cells in communication with the routing layer for permitting exchange of data packets between the array of cells and the routing layer. Each cell includes a memory for receiving a data packet from the routing layer. The routing layer includes a controller to control release of a data packet toward a cell of the array at least in part on a basis of a degree of occupancy of the memory in the cell.

Abstract: A switch fabric implemented on a chip includes an array of cells and an I/O interface in communication with the array of cells for permitting exchange of data packets between the array of cells and components external to the array of cells. Each cell communicates with at least one other cell of the array, permitting an exchange of data packets between the cells of the array and an exchange of control information between the cells of the array. Each cell is operative to control transmission of data packets to other cells of the array at least in part on a basis of the control information. The control information is thus used to regulate the flow of data packets between cells.

Abstract: A switch fabric implemented on a chip includes an array of cells and an I/O interface in communication with the array of cells for permitting exchange of data packets between said array of cells and components external to said array of cells. Each cell communicates with at least one other cell of the array, thereby permitting an exchange of data packets to take place between the cells of the array. Each cell includes a memory for receiving a data packet from another cell of the array as well as a control entity to control release of a data packet toward a selected destination cell of the array at least in part on a basis of a degree of occupancy of the memory in the destination cell. In this way, scheduling is distributed amongst the cells of the switch fabric.

Abstract: A data processing system containing a monolithic network of cells with sufficient redundancy provided through direct logical replacement of defective cells by spare cells to allow a large monolithic array of cells without uncorrectable defects to be organized, where the cells have a variety of useful properties. The data processing system according to the present invention overcomes the chip-size limit and off-chip connection bottlenecks of chip-based architectures, the von Neumann bottleneck of uniprocessor architectures, the memory and I/O bottlenecks of parallel processing architectures, and the input bandwidth bottleneck of high-resolution displays, and supports integration of up to an entire massively parallel data processing system into a single monolithic entity.

Abstract: A data processing system containing a monolithic network of cells with sufficient redundancy provided through direct logical replacement of defective cells by spare cells to allow a large monolithic array of cells without uncorrectable defects to be organized, where the cells have a variety of useful properties. The data processing system according to the present invention overcomes the chip-size limit and off-chip connection bottlenecks of chip-based architectures, the von Neumann bottleneck of uniprocessor architectures, the memory and I/O bottlenecks of parallel processing architectures, and the input bandwidth bottleneck of high-resolution displays, and supports integration of up to an entire massively parallel data processing system into a single monolithic entity.

Abstract: A data processing system containing a monolithic network of cells with sufficient redundancy provided through direct logical replacement of defective cells by spare cells to allow a large monolithic array of cells without uncorrectable defects to by organized, where the cells have a variety of useful properties. The data processing system according to the present invention overcomes the chip-size limit and off-chip connection bottlenecks of chip-based architectures, the von Neumann bottleneck of uniprocessor architectures, the memory and I/O bottlenecks of parallel processing architectures, and the input bandwidth bottleneck of high-resolution displays, and supports integration of up to an entire massively parallel data processing system into single monolithic entity.

Abstract: A data processing system containing a monolithic network of cells with sufficient redundancy provided through direct logical replacement of defective cells by spare cells to allow a large monolithic array of cells without uncorrectable defects to be organized, where the cells have a variety of useful properties. The data processing system according to the present invention overcomes the chip-size limit and off-chip connection bottlenecks of chip-based architectures, the von Neumann bottleneck of uniprocessor architectures, the memory and I/O bottlenecks of parallel processing architectures, and the input bandwidth bottleneck of high-resolution displays, and supports integration of up to an entire massively parallel data processing system into a single monolithic entity.

Abstract: A massively parallel data processing system consisting of an array of closely spaced cells where each cell has direct output means as well as means for processing, memory and input. The data processing system according to the present invention overcomes the von Neumann bottleneck of uniprocessor architectures, the I/O and memory bottlenecks that plague parallel processors, and the input bandwidth bottleneck of high-resolution displays.

Abstract: A data processing system containing a monolithic network of cells with sufficient redundancy provided through direct logical replacement of defective cells by spare cells to allow a large monolithic array of cells without uncorrectable defects to be organized, where the cells have a variety of useful properties. The data processing system according to the present invention overcomes the chip-size limit and off-chip connection bottlenecks of chip-based architectures, the von Neumann bottleneck of uniprocessor architectures, the memory and I/O bottlenecks of parallel processing architectures, and the input bandwidth bottleneck of high-resolution displays, and supports integration of up to an entire massively parallel data processing system into a single monolithic entity.