Abstract:
An arrangement comprising a node having a node input configured to receive a plurality of transactions intended for a plurality of different targets, said node having a plurality of node outputs; at least one target, said at least one target comprising an input configured to receive a respective output of said node; an output, wherein said node is configured to direct said transactions to one of said at least one target and said output depending on if said transactions are intended for said at least one target or at least one different target.
Abstract:
A method of perform transactions in a communication network, wherein information is exchanged between Intellectual Property (IP) cores. The information is transported in packets (P) including a header (HD) for transporting control information, and one or more payloads (PL) transporting content. The method provide the use of a versatile packet format (HF1-HF17) adapted to transport different traffic patterns generated by Intellectual Property (IP) cores using different protocols to ensure a simple interoperability between the Intellectual Property (IP) cores, providing also the configurability of the granularity arbitration process in order to correct crossing the routers in the communication network.
Abstract:
A method for making an on-chip interconnect for conveying between a set of initiators (10) and a set of targets (20) a traffic organized in a number of classes of service (ABC, SRT, HRT, LC) envisages: - associating to the traffic, starting from the aforesaid initiators (10) respective priority values representing the classes of service; - propagating the priority values towards the points (21, 22) of the network, where an arbitration is performed between two classes of service of the aforesaid traffic (ABC, SRT, HRT, LC); and - providing in said points the arbitration as a function of the aforesaid priority values.