| 摘要 |
<p>A computer system package minimises latency between many high performance processors and a large amount of shared memory. Wire length, latency and skew are minimised while supporting a very fast cycle time and high bandwidth by stacking edge connected modules (ECMs), which are the basic building blocks. The ECMs 20 are characterised by signal input/output (I/O) pads on three edges, the two opposing edges being referred to as the inside connector edges and the third edge being referred to as the global connector edge. The ECMs support multiple processors per module, a plurality of basic storage modules (BSMs) per module, and portions of request and response switches per module. A plurality of processor ECMs and request switch ECMs are stacked in a first stack, and a plurality of BSM ECMs and response switch ECMs are stacked in a second stack. The two stacks are arranged adjacent one another with the request switch ECMs above the processor ECMs and the response switch ECMs below the BSM ECMs so that the response switch ECMs are adjacent the processor ECMs and the request ECMs are adjacent the BSM ECMs. The data flow from a processor to addressed memory and back to the processor is unidirectional about a loop defined by connections on the inside edges of the ECMs. The stacks form a subsystem which may be used as a stand-alone system or interconnected with other subsystems via the global connectors connected to the third edges of the ECMs. The global connectors allow the processors of one subsystem to be connected to the BSMs of another subsystem. <IMAGE></p> |
