Interconnect system to support the execution of instruction sequences by a plurality of partitionable engines
View Patent ↗A global interconnect system. The global interconnect system includes a plurality of resources having data for supporting the execution of multiple code sequences and a plurality of engines for implementing the execution of the multiple code sequences. A plurality of resource consumers are within each of the plurality of engines. A global interconnect structure is coupled to the plurality of resource consumers and coupled to the plurality of resources to enable data access and execution of the multiple code sequences, wherein the resource consumers access the resources through a per cycle utilization of the global interconnect structure.
1. A multicore processor having an interconnect system, the system comprising:
a plurality of resources to store data to support execution of multiple code sequences;
a plurality of partitionable engines to implement the execution of the multiple code sequences;
a plurality of resource consumers where at least one resource consumer of the plurality of resource consumers is located within each of the plurality of partitionable engines;
an interconnect structure for coupling the plurality of resource consumers with the plurality of resources to access the data and to execute the multiple code sequences, wherein the interconnect structure includes a routing matrix of transmission lines that allows each of the plurality of resource consumers to directly access data from any storage location within the plurality of resources based on reservations of the transmission lines per request for access to data from a storage location within the plurality of resources, and wherein the interconnect structure is a segmented interconnect structure,
wherein the plurality of resources are contended for by the plurality of resource consumers and the plurality of resource consumers send fetch requests to a reservation adder, which adds a number of fetch requests to produce a sum, and a threshold limiter checks whether the sum is under a threshold to determine that the plurality of resource consumers is able to reserve a corresponding contested resource.
2. The multi core processor of claim 1 , wherein the plurality of resource consumers are selected from the group consisting of: execution units and address calculation units of the partitionable engines.
3. The multicore processor of claim 1 , wherein the plurality of resources comprise memory fragments.
4. The multicore processor of claim 1 , wherein the plurality of resources comprise register file segments.
5. The multicore processor of claim 1 , wherein the plurality of resources are contended for by the plurality of resource consumers as the plurality of resource consumers attempt to fetch resources necessary to support consumption.
6. The multi core processor of claim 1 , wherein the segmented interconnect structure comprises a plurality of global segmented buses and a plurality of locally connected engine-to-engine buses.
7. The multi core processor of claim 6 , wherein a segmentation of the segmented interconnect structure is implemented using tri-state buffers.
8. The multi core processor of claim 7 , wherein the plurality of resources comprise segments of the segmented interconnect structure.
9. A microprocessor, comprising
a plurality of resources to store data to support the execution of multiple code sequences;
a plurality of partitionable engines to implement the execution of the multiple code sequences;
a plurality of resource consumers where at least one resource consumer of the plurality of resource consumers is located within each of the plurality of partitionable engines;
an interconnect structure for coupling the plurality of resource consumers with the plurality of resources to access the data and to execute the multiple code sequences, wherein the interconnect structure includes a routing matrix of transmission lines that allows each of the plurality of resource consumers to directly access data from any storage location within the plurality of resources based on reservations of the transmission lines per request for access to data from a storage location within the plurality of resources, and wherein the interconnect structure is a segmented interconnect structure operable to implement a fetch model interconnect structure,
wherein the plurality of resources are contended for by the plurality of resource consumers and the plurality of resource consumers send fetch requests to a reservation adder, which adds a number of fetch requests to produce a sum, and a threshold limiter checks whether the sum is under a threshold to determine that the plurality of resource consumers is able to reserve a corresponding contested resource.
10. The microprocessor of claim 9 , wherein the plurality of resource consumers are selected from the group consisting of: execution units and address calculation units of the partitionable engines.
11. The microprocessor of claim 9 , wherein the plurality of resources are contended for by the plurality of resource consumers as the plurality of resource consumers attempt to fetch resources necessary to support consumption.
12. The microprocessor of claim 9 , wherein the segmented interconnect structure comprises a plurality of global segmented buses and a plurality of locally connected engine to engine buses.
13. The microprocessor of claim 12 , wherein a segmentation of the segmented interconnect structure is implemented using tri-state buffers.
14. The microprocessor of claim 13 , wherein the plurality of resources comprise segments of the segmented interconnect structure.
15. The microprocessor of claim 13 , wherein the tri-state buffers comprise one-directional tri-state buffers and bidirectional tri-state buffers.
16. A computer system having a microprocessor coupled to a computer readable memory, wherein the microprocessor comprises:
a plurality of resources to store data to support the execution of multiple code sequences;
a plurality of partitionable engines to implement the execution of the multiple code sequences;
a plurality of resource consumers where at least one resource consumer of the plurality of resource consumers is located within each of the plurality of partitionable engines;
an interconnect structure for coupling the plurality of resource consumers with the plurality of resources to access the data and to execute the multiple code sequences, wherein the interconnect structure includes a routing matrix of transmission lines that allows each of the plurality of resource consumers to directly access data from any storage location within the plurality of resources based on reservations of the transmission lines per request for access to data from a storage location within the plurality of resources, and wherein the interconnect structure is a segmented interconnect structure comprising a plurality of global segmented buses and a plurality of locally connected buses,
wherein the plurality of resources are contended for by the plurality of resource consumers and the plurality of resource consumers send fetch requests to a reservation adder, which adds a number of fetch requests to produce a sum, and a threshold limiter checks whether the sum is under a threshold to determine that the plurality of resource consumers is able to reserve a corresponding contested resource.