Computationally-networked unified data bus
View Patent ↗Embodiments of the present invention provide a computationally-networked unified data bus for a multi-processing domain architecture. Specifically, in a typical embodiment, a unified data bus is provided. A first data bus adapter (e.g., a node) is coupled to the unified data bus (e.g., a link), and a first processing domain is coupled to the first data bus adapter. In general, the first data bus adapter encapsulates, translates, and interprets data communicated between the unified data bus and the first processing domain. In addition, a second data bus adapter (e.g., a node) is coupled to the unified data bus and a second processing domain is coupled to the second data bus adapter. Similar to the first data bus adapter, the second data bus adapter encapsulates, translates, and interprets data communicated between the unified data bus and the second processing domain. Under these embodiments, the first processing domain and the second processing domain can each comprise at least one element selected from a group consisting of: memory input/outputs (I/Os), cache, heterogeneous data buses, and processors. Moreover, the first processing domain and the second processing domain can be selected from a group consisting of a heterogeneous processing domain and a hybrid processing domain.
1. A computationally-networked unified data bus for a multi-core architecture, comprising:
a unified data bus;
a first data bus adapter coupled to the unified data bus;
a first core, comprising a first customer data bus, coupled to the first data bus adapter, the first data bus adapter configured to encapsulate, translate, and interpret data communicated to the unified data bus from the first core from a first format of the first core into a common format of the unified data bus and to the first core from the unified data bus from the common format to the first format for delegation of processing to a second core, wherein the first data bus adapter is specific to the first core and wherein all data communicated to the first core and all data communicated from the first core pass through the first data bus adapter;
a second data bus adapter coupled to the unified data bus; and
the second core, comprising a second customer data bus, coupled to the second data bus adapter, the second data bus adapter configured to encapsulate, translate, and interpret data communicated to the unified data bus from the second core from a second format of the second core into the common format of the unified data bus and to the second core from the unified data bus from the common format to the second format for delegation of processing to the first core, wherein the second data bus adapter is specific to the second core and wherein all data communicated to the second core and all data communicated from the second core pass through the second data bus adapter;
wherein the first core and the second core are hybrid cores.
2. The computationally-networked unified data bus of claim 1 , the first core and the second core each comprising at least one element selected from a group consisting of: memory input/outputs (I/Os), cache, heterogeneous data buses, and processors.
3. The computationally-networked unified data bus of claim 1 , the first bus adapter and the second data bus adapter comprising nodes.
4. The computationally-networked unified data bus of claim 1 , the data comprising an entity that includes memory addressing information and content to enable out-of-immediate memory virtualization, instruction translation, delegated instruction interpretation, faster cache virtualization, and dynamic bus management.
5. The computationally-networked unified data bus of claim 4 , the data being packaged with a predefined format.
6. A computationally-networked unified data bus for a multi-processing domain architecture, comprising:
a unified data bus;
a first data bus adapter coupled to the unified data bus;
a first processing domain, comprising a first collaborative processing mode and a first customer data bus, the first customer data bus coupled to the first data bus adapter, the first data bus adapter configured to encapsulate, translate, and interpret data communicated to the unified data bus from the first processing domain from the first collaborative processing mode of the first processing domain into a common format of the unified data bus and to the first processing domain from the unified data bus from the common format to the first collaborative processing mode, wherein the first data bus adapter is specific to the first processing domain and wherein all data communicated to the first processing domain and all data communicated from the first processing domain pass through the first data bus adapter;
the second data bus adapter coupled to the unified data bus; and
a second processing domain, comprising a second collaborative processing mode and a second customer data bus, the second processing domain coupled to the second data bus adapter, the second data bus adapter configured to encapsulate, translate, and interpret data communicated to the unified data bus from the second processing domain from the second collaborative processing mode of the second processing domain into a common format of the unified data bus and to the second processing domain from the unified data bus from the common format to the second collaborative processing mode, wherein the second data bus adapter is specific to the second processing domain and wherein all data communicated to the second processing domain and all data communicated from the second processing domain pass through the second data bus adapter;
wherein the first processing domain and the second processing domain are hybrid processing domains configured to collaborate with one another across the unified data bus via the first collaborative processing mode and the second collaborative processing mode.
7. The computationally-networked unified data bus of claim 6 , the first processing domain and the second processing domain each comprising at least one element selected from a group consisting of: memory input/outputs (I/Os), cache, heterogeneous data buses, and processors.
8. The computationally-networked unified data bus of claim 6 , the first bus adapter and the second data bus adapter comprising nodes, the nodes connected to one another by a link, the link comprising one or more of the following functions: connection configuration, real-time bus optimization, and routing.
9. The computationally-networked unified data bus of claim 6 , the data comprising an entity that includes memory addressing information and content to enable out-of-immediate memory virtualization, instruction translation, delegated instruction interpretation, faster cache virtualization, and dynamic bus management.
10. The computationally-networked unified data bus of claim 9 , the data being packaged with a predefined format.
11. A method for producing a computationally-networked unified data bus for a multi-processing domain architecture, comprising:
providing a unified data bus;
coupling a first data bus adapter coupled to the unified data bus;
coupling a first processing domain, comprising a first collaborative processing mode and a first custom data bus, coupled to the first data bus adapter, the first data bus adapter configured to encapsulate, translate, and interpret data communicated to the unified data bus from the first processing domain from the first collaborative processing mode of the first processing domain into a common format of the unified data bus and to the first processing domain from the unified data bus from the common format to the first collaborative processing mode, for delegation of processing to a second processing domain, wherein the first data bus adapter is specific to the first processing domain and wherein all data communicated to the first processing domain and all data communicated from the first processing domain pass through the first data bus adapter;
coupling a second data bus adapter coupled to the unified data bus; and
coupling a second processing domain, comprising a second collaborative processing mode and a second custom data bus, coupled to the second data bus adapter, the second data bus adapter configured to encapsulate, translate, and interpret data communicated to the unified data bus the second processing domain from the second collaborative processing mode of the second processing domain into a common format of the unified data bus and to the second processing domain from the unified data bus from the common format to the second collaborative processing mode, for delegation of processing to the first processing domain, wherein the second data bus adapter is specific to the second processing domain and wherein all data communicated to the second processing domain and all data communicated from the second processing domain pass through the second data bus adapter;
wherein the first processing domain and the second processing domain are hybrid processing domains configured to collaborate with one another across the unified data bus via the first collaborative processing mode and the second collaborative processing mode.
12. The method of claim 11 , the first processing domain and the second processing domain each comprising at least one element selected from a group consisting of: memory input/outputs (I/Os), cache, heterogeneous data buses, and processors.
13. The method of claim 11 , the first bus adapter and the second data bus adapter comprising nodes, the nodes connected to one another by a link.
14. The method of claim 13 , wherein the nodes comprises one or more of the following functions: connection configuration, real-time bus optimization, and routing.
15. The method of claim 11 , the data comprising an entity that includes memory addressing information and content to enable out-of-immediate memory virtualization, instruction translation, delegated instruction interpretation, faster cache virtualization, and dynamic bus management.
16. The method of claim 15 , the data being packaged with a predefined format.