Vlsi Technology, Inc.

Freescale Semiconductor,inc.

Ned D Garinger

Gary A Walker

Martin L Dorr

Mark W Naumann

Peter H Sorrells

Tein-Yow Yu

Nicholas Julian Richardson

Joseph Murray

Joseph Alan Thomsen

Gabriel R Munguia

Jeffery M Michelsen

Kenneth P Caviasca

James B Nolan

W Henry Potts

Pratiksh Parikh

An OCN for integrated processing elements including a network with multiple ports and multiple port interfaces. The ports and the port interfaces conform to a consistent port protocol. Each port interface converts information between bus transactions of a corresponding processing element and network packets and exchanges network packets with other port interfaces. Each port includes an arbitration interface and a data interface and the network includes an interconnect and an arbiter. The interconnect includes selectable data paths between the ports for packet datum transfer. A port source interface submits transaction requests and provides packet datums upon receiving an acknowledgement. A port destination interface receives packet datums via available input buffers. Each transaction request includes a transaction size and a destination port address. The arbiter receives transaction requests, arbitrates among transaction requests, provides acknowledgements and controls the interconnect to select data paths between sources and destinations.

On chip network

A network that maximizes interconnect utilization between integrated processing elements, including ports, an interconnect, port interfaces, and an arbiter. Each port includes arbitration and data interfaces. The interconnect includes selectable data paths between the ports for packet datum transfer. Each port interface includes processing, source and destination interfaces. The source interface submits transaction requests and provides packet datums upon receiving an acknowledgement. The destination interface receives packet datums via a number of available input buffers. Each transaction request includes a transaction size, a packet priority, and a destination port address. The arbiter includes a request queue and a buffer counter for each port and a datum counter for each acknowledged transaction. The arbiter arbitrates among transaction requests based on a selected arbitration scheme, destination buffer availability, data path availability, and priority, and uses the packet datum counters, arbitration latency and data path latency to minimize dead cycles in the interconnect.

On chip network that maximizes interconnect utilization between processing elements

An OCN with independent logical and physical layers for enabling communication among integrated processing elements, including ports, bus gaskets and a physical layer interface. Each bus gasket includes a processor element interface and a port interface. Each processor element interface of at least two bus gaskets operates according to a first logical layer protocol. Each port interface operates according to a consistent port interface protocol by sending transaction requests and receiving acknowledgements and by sending and receiving packet datums via the corresponding port. The physical layer interface transfers packets between the ports and includes an arbiter and an interconnect coupled to each port. Additional bus gaskets may be added that operate according to a second logical layer protocol which may or may not be compatible with the first. Any bus gasket may be added that is configured to communicate using multiple logical layer protocols.

On chip network with independent logical and physical layers

A scalable network for supporting an application using processing elements including ports, an interconnect, port interfaces, and an arbiter. Each port conforms to a consistent port interface protocol regardless of number of ports, frequency of operation, maximum datum width or data path concurrency. The interconnect has a scalable maximum datum width and a scalable data path concurrency, and includes selectable data paths between any two ports to enable transfer of datums between the ports. Each port interface formulates packets for transmission and receives packets via the corresponding port and the interconnect, where each packet includes one or more datums. The arbiter controls packet transfer via the interconnect between source and destination ports. The interconnect has a scalable data path concurrency. Pipeline stages may be added to support a selected clock frequency. The OCN may be a component library including bus gasket, interconnect and arbiter components.

Scalable on chip network

A network with memory device address decoding that enables communication among integrated processing elements, including a network, a processing element and a bus gasket. The network transfers packets between multiple ports, where each port conforms to a consistent port interface protocol. The processing element includes a bus and a memory device programmed with the address of each port, so that a transaction on the bus indicating another port is decoded by the memory device. The bus gasket includes a bus interface that generates packets and a port interface that sends and receives the packets according to the consistent port interface protocol and that uses the decoded address as a destination port address. The memory device may be implemented in any desired manner, such as a memory management unit (MMU) or a direct memory access (DMA) device.

On chip network with memory device address decoding

Level trigger mode interrupts are converted to edge trigger mode interrupts in a computer system. A circuit detects the occurrence of a level trigger mode interrupt request, and asserts an edge trigger mode interrupt request output. The edge trigger mode interrupt request remains asserted until an End of Interrupt input is asserted, indicating that the CPU has completed servicing the prior interrupt. The edge trigger mode interrupt request is then deasserted.

Circuit and method for converting interrupt signals from level trigger

An apparatus and method for minimizing bus traffic by combining write operations is disclosed. The present invention detects the occurrence of consecutive byte updates to a common 32-bit block. This is accomplished by using comparators to examine the addresses of consecutive write operations. If it is determined that the consecutive write operations are indeed to a common 32-bit block, they are combined. The address of the next write operation is also, similarly checked. All of the writes into that particular block are combined in a write combine register. The contents of this register is then transferred to a write buffer. When bus access is granted, the combined byte updates stored in the write buffer are issued in a single memory write cycle to the bus, thereby minimizing the number of write cycles actually required to transfer the data.

Combined consective byte update buffer

A microchip circuit for use in and method for use with PC/AT microprocessor environments enables both local memory and ISA bus memory to be refreshed while reducing CPU overhead time entailed in a performing refresh by providing decoupled refresh cycles for the local memory and the ISA memory.

Decoupled refresh on local and system busses in a PC/at or similar

An output pad for an integrated circuit includes circuitry to align the output with an on-chip clock signal, and to compensate the output such that it remains coincident with the on-chip clock signal even when changes occur in power supply voltage, manufacturing process and temperature. This output pad has a closed-loop feedback circuit which controls the delay of the output signal through a variable delay element. The loop adjust the delay until the clock edge of the on-chip clock signal is coincident with the output signal within a defined tolerance. The output pad self-compensates with every clock cycle, which is many times faster than any induced variation.

Self-compensating output pad for an integrated circuit and method

An output pad for an integrated circuit includes circuitry to fix the output delay to a constant value even when changes occur in power supply voltage, manufacturing process and temperature. This output pad has a closed-loop feedback circuit which controls the delay of the output signal through an Adjustable Output Driver. A Fixed Delay Element provides a reference delay that does not change with induced variations. The loop adjusts the delay of the Adjustable Output Driver until the edge of the signal delayed by the Adjustable Output Driver is coincident with the edge of the signal delayed by the Fixed Delay Element within a defined tolerance. In a second embodiment the output pad has a Variable Delay Element with programmable delay values that remain constant over induced variations. In a third embodiment the Fixed Delay Element is deleted, and the delayed signal through the Adjustable Output Driver is compared to and therefore synchronized with a Reference Signal, which may be provided on-chip or from a source external to the chip.

Delay-compensated output pad for an integrated circuit and method

A compensated digital delay semiconductor device is disclosed which uses two identical chains of delay elements. The first chain is the Reference Chain, which is driven by a crystal-controlled digital clock input. The second chain is the Input Signal Delay Chain, which is the delay path for the signal of interest. These two chains are located in physical proximity on the semiconductor die so that variations in the manufacturing process, temperature and power supply affect each chain the same. Circuitry monitors the delay performance of the Reference Chain, and dynamically changes the output tap of the Input Signal Delay Chain when a change in performance is detected on the Reference Chain, thereby compensating the delay of the device. This approach provides precise delays which are constant.

Compensated digital delay semiconductor device with selectable output

A self-compensated digital delay semiconductor device is disclosed which uses two identical chains of delay elements. The first chain is the Reference Chain, which is driven by a crystal-controlled digital clock input. The second chain is the Input Signal Delay Chain, which is the delay path for the signal of interest. These two chains are located in physical proximity on the semiconductor die so that variations in manufacturing process, temperature and power supply affect each chain the same. Each of these delay chains is comprised of a series of variable delay elements which are digitally controlled by Monitor Logic, which measures the delay performance of the Reference Chain, and dynamically adjusts the delay of the variable delay elements as induced variations are induced, thereby compensating the delay of the device. Any one of these precise delays can be routed to the output by driving a tap select multiplexer to select the delay of interest. This approach provides precise delays which are constant within a tight tolerance.

Self-compensating digital delay semiconductor device with selectable

A circuit generates a system clock signal. On a first input of the circuit a first oscillating signal is placed. On a second input, a second oscillating signal may be placed. Clock sense logic is connected to the second input. The clock sense logic detects whether the second oscillation signal is present on the second input. When the second oscillating signal is not present on the second input, the first oscillating signal is selected to be used to generate the system clock. When the second oscillating signal is present on the second input, the second oscillating signal is selected to be used to generate the system clock. The selected oscillating signal is divided to produce the system clock signal. A first frequency divider divides the selected oscillating signal by a first amount. In parallel, a second frequency divider divides the selected oscillating signal by a second amount. A selector, for example, a multiplexor, selects output from either the first frequency divider or the second frequency divider as the system clock signal. When the first oscillating signal is being used to generate the system clock, the second input may be used to control the selection of frequency dividers.

Growing community of inventors

Tempe, AZ, United States of America

Ned D Garinger