Abstract:
An integrated circuit device (11) has an address and data path (15) interconnecting a CPU (12) with modules (14) the modules having event logic (8) to generate an event request packet having a destination address and the CPU decoding the packet to selectively respond to the request of the packet depending on the priority of the event.
Abstract:
A chip (11) with an address and data path (15) interconnecting at least one CPU (12) with another module (14) and an external communication port (30), the CPU generating event packets and memory access packets and the module (14) generating event packets, the packets being distributed in parallel format on the path 15 and the external communication port (30) including circuitry to reduce the parallel format of each packet to a more serial format for off-chip communication.
Abstract:
An integrated circuit chip (11) has an address and data path (15) interconnecting at least one CPU (12) with a different module (15), the CPU and module each having event logic (8,44) to generate event request packets of two types, each having a destination address, one type being a control packet to which a destination device must respond on receipt and the other type including a priority indicator such that the destination device can selectively respond depending on the priority detected.
Abstract:
An integrated circuit device (11) with an address and data path (15) interconnects a CPU (12) with at least one module (14) and a memory interface (32), the module (14) having circuitry (8) to generate an event request packet and the CPU having event logic (44) to decode the packet as well as circuitry to generate addressed memory access packets, the same address and data path (15) being used for the distribution of event request packets and memory access packets.
Abstract:
A computer system comprises interconnected chips (11) each having an on-chip address and data path (15), a CPU (12) arranged to send and receive addressed bit packets on the path and modules (14) with event circuitry (8) to generate addressed event request packets for distribution on the path (15), the two chips being interconnected through respective external ports (30) so that locations on each of the paths (15) form part of a common address space accessible from either chip.