Abstract: A system transfers a data stream including data packets separated by non-packet words from a first clock domain to a second clock domain. It includes an elasticity buffer into which the data stream is written in a cyclic sequence under the control of the clock frequency in the first clock domain and from which the data stream is read out in a cyclic sequence under the control of the clock frequency in the second domain. The two sequences are monitored to provide an anticipatory signal indicating that the reading sequence approaches proximity to the writing sequence. A non-packet word is inserted into the data stream in the first domain. In the second clock domain the existence of the inserted non-packet word is detected and the buffer is caused to advance the reading cycle thereby to discard the said inserted non-packet word.

Abstract: A program tool automatically generating interconnect logic for a system-on-a-chip is based on a library of operational cores and on a architecture which requires all data exchange between cores to proceed via shared memory, which may be ‘off-chip’. The architecture includes a data aggregation technique for access to memory with successive levels of arbitration.

Abstract: An architecture for a system on a chip wherein functional cores have wrappers for compatibility with a common bus system and the bus system includes aggregators for bus transactions of different speeds and/or bus widths.

Abstract: A stackable network unit which can form a ring with other units has a master mode in which it can place packets on the ring, so that they can be ultimately forwarded from other units, and a repeat mode in which it can make a request for transfer of mastership of the ring, inserting bits in a header of an arbitration packet. If a master unit has completed the transmission of a packet and, preferably, after the required inter-packet gap has elapsed, it has not yet received an arbitration header to permit the making of an arbitration decision it can transmit a subsequent packet. The header information of this packet indicates to the other units that this subsequent packet is part of a burst of packets and the other units should not set requests in the header of this packet because this packet would not be used for arbitration.

Abstract: A method and apparatus for refresh command operations on an SDRAM that avoids use of refresh commands requiring all banks of the SDRAM to be idle. Burst operation establishes command sequences that include Nop command intervals. Some of these Nop intervals are used to perform operations on a bank other than the one under access for the burst that provide a refresh. ACTIVE followed by PRECHARGE commands are inserted into the command intervals addressed to a refresh address. The refresh addresses are generated externally of the SDRAM and provided to a multiplexer that sequences them with the data addresses. A secondary timer checks that required refresh has occurred and prioritizes the refresh addresses over data addresses in the multiplexer in the event that a refresh has not been completed shortly before a maximum refresh interval.

Abstract: Burst operation establishes command sequences that include Nop command intervals Some of these Nop intervals are used to perform operations on a bank other than the one under access for the burst that provide a refresh. ACTIVE followed by PRECHARGE commands are inserted into the command intervals addressed to a refresh address The refresh addresses are generated externally of the SDRAM and provided to a multiplexer that sequences them with the data addresses.

Abstract: A data bus system in which a read or write transaction includes an identification of the initiator of the transaction and optionally an identification of the transaction as a number in a cyclic progression and optionally a request for an acknowledgement.

Abstract: An architecture for a system on a chip wherein functional cores have wrappers for compatibility with a common bus system and the bus system includes aggregators for bus transactions of different speeds and/or bus widths