A memory device including an interface circuit for data conversion according to different endian formats includes an interface circuit that performs data conversion with hardware in a data transfer path inside the memory device in accordance with a memory bank, a processing element (PE), and an endian format of a host device. The interface circuit is (i) between a memory physical layer interface (PHY) region and a serializer/deserializer (SERDES) region, (ii) between the SERDES region and the memory bank or the PE, (iii) between the SERDES region and a bank group input/output line coupled to a bank group including a number of memory banks, and (iv) between the PE and bank local input/output lines coupled to the memory bank.

A network comprising interconnected first and second processors, each processor comprising one or more of: multiple processing units arranged on a chip configured to execute program code; an on-chip interconnect comprising groups of exchange paths connected to receive data from corresponding groups of the processing units; external interfaces configured to communicate data off-chip as packets, each having a destination address, external interfaces of the first and second processors being connected by an external link; multiple exchange blocks, each connected to groups of the exchange paths; a routing bus configured to route packets between the exchange blocks and the external interfaces. Processing units of the first processor generate off-chip packets such that the group of processing units serviced by the first exchange block on the first processor address off-chip packets to the group of processing units on the second processor serviced by the corresponding first exchange block of the second processor.

Disclosed is an Avalon-to-Axi4 bus conversion method, including: in case that an Avalon bus is an Avalon_st bus, receiving Avalon_st bus data, performing a logical process on the received Avalon_st bus data, and then outputting corresponding Axi4_st bus data; and in case that the Avalon bus is an Avalon_mm bus, receiving a signal transmitted by each channel of the Avalon_mm bus, framing and storing the signal in asynchronous First Input First Output (FIFO), and in case that a device corresponding to an Axi4 bus is ready, reading the signal from the asynchronous FIFO, and outputting the signal to a corresponding channel of the Axi4 bus according to a timing relationship of the Axi4 bus.

A hardware direct memory access controller including an input port configured to receive data from an electronic device for direct memory access transfer, an output port configured to provide data for direct memory access, and processing circuitry is disclosed. The processing circuitry is configured to receive data comprising a header and payload from the electronic device via the input port, parse the header to determine data parameters including a transaction length and an input data format, and select a target destination for the data based at least in part on the data parameters. The processing circuitry is also configured to allocate memory within the target destination based at least in part on the transaction length, and to format the payload for direct memory access based at least in part on the data parameters, and to transfer the formatted payload for storage within the allocated memory within the target destination via the output port using direct memory access.

An on-chip cache is described which receives memory requests and in the event of a cache miss, the cache generates memory requests to a lower level in the memory hierarchy (e.g. to a lower level cache or an external memory). Data returned to the on-chip cache in response to the generated memory requests may be received out-of-order. An instruction scheduler in the on-chip cache stores pending received memory requests and effects the re-ordering by selecting a sequence of pending memory requests for execution such that pending requests relating to an identical cache line are executed in age order and pending requests relating to different cache lines are executed in an order dependent upon when data relating to the different cache lines is returned. The memory requests which are received may be received from another, lower level on-chip cache or from registers.

A data processing system comprising a plurality of data inputs and of data outputs for processing input data and providing processed data to a data output. The system comprises a plurality of data processing hardware units, each being configured to process data within a predetermined latency and according to a data processing task of a predetermined type. The system further comprises a memory for storing a predetermined latency for each of the data processing hardware units and a controller configured to determine a type of a data processing task to be executed as a function of a source of data to be processed or of a destination of processed data and further configured to select one data processing hardware unit as a function of the determined type of the task to be executed and of latency constraints associated with the task to be executed.

A system includes an input/output adapter operable to receive a plurality of packets in a single clock cycle. The system includes a controller operatively connected to the input/output adapter. The controller is operable to receive a first packet at a data link layer and determine a state of a first output indicator to maintain packet ordering. Based on determining that a first receiver formatting interface is selected by the first output indicator, the controller performs an alignment adjustment and output of the first packet by the first receiver formatting interface. Based on determining that a second receiver formatting interface is selected by the first output indicator, the controller performs the alignment adjustment and output of the first packet by the second receiver formatting interface.

Disclosed embodiments relate, generally, to interfacing serial communication interfaces of a first device with a parallel communication interface of a second device. A first group of two or more serial communication interfaces and an interfacing logic may be provided. The interfacing logic may form second encoded data blocks by arranging the data elements of the first encoded data blocks such that data elements within a same data element position of respective second encoded data blocks represent a given one of the symbols, and provide the second encoded data blocks to a number of serial communication interfaces coupled to a parallel communication interface of another device. An interfacing logic may additionally or alternatively be configured to receive, from a second group of two or more serial communication interfaces, received encoded data blocks representing received symbols.

An apparatus to facilitate memory barriers is disclosed. The apparatus comprises an interconnect, a device memory, a plurality of processing resources, coupled to the device memory, to execute a plurality of execution threads as memory data producers and memory data consumers to a device memory and a system memory and fence hardware to generate fence operations to enforce data ordering on memory operations issued to the device memory and a system memory coupled via the interconnect.

A memory device including an interface circuit for data conversion according to different endian formats includes an interface circuit that performs data conversion with hardware in a data transfer path inside the memory device in accordance with a memory bank, a processing element (PE), and an endian format of a host device. The interface circuit is (i) between a memory physical layer interface (PHY) region and a serializer/deserializer (SERDES) region, (ii) between the SERDES region and the memory bank or the PE, (iii) between the SERDES region and a bank group input/output line coupled to a bank group including a number of memory banks, and (iv) between the PE and bank local input/output lines coupled to the memory bank.