Dynamically composable computing system, a data center, and method for dynamically composing a computing system

1. A system, comprising:

a computing fabric including a plurality of different disaggregated computing hardware resources having respective hardware characteristics;

a resource manager having access to the respective hardware characteristics of the different disaggregated computing hardware resources and configured to assemble a composite computing node by selecting one or more disaggregated computing hardware resources with respective hardware characteristics meeting requirements of an application to he executed on the composite computing node; and

an orchestrator configured to schedule the application using the assembled composite computing node,

wherein the orchestrator is configured to determine, according to the requirements of the application, a number of memory and/or storage tiers of the composite computing node, wherein different memory and/or storage tiers are associated with respective different hardware resource characteristics, and wherein the resource manager is configured to associate one or more memory or storage hardware resources meeting the respective hardware resource characteristic with a respective memory or storage tier and map a respective address range to a respective memory or storage tier.

2. The system of claim 1 , wherein respective hardware characteristics comprise processing performance, power consumption, bandwidth, latency, capacity, underlying technology, or cost.

3. The system of claim 1 , wherein the resource manager includes storage configured to store metadata indicative of the hardware characteristics associated with the plurality of different distributed disaggregated hardware resources.

4. The system of claim 1 , wherein the different distributed computing hardware resources comprise different types of computer memory or storage.

5. The system of claim 1 , further comprising

at least one compute sled including a set of local computing hardware resources,

wherein the resource manager is configured to connect the compute sled to the selected disaggregated computing hardware resources through the computing fabric.

6. The system of claim 5 , wherein the local computing hardware resources of the compute sled comprise one or more local CPUs and one or more local memory modules.

7. The system of claim 5 , wherein the compute sled includes a set of control registers and wherein the resource manager is configured to specify the selected disaggregated computing hardware by writing to the control registers.

8. The system of claim 7 , wherein the set of control registers is indicative of at least one of a number of memory or storage tiers, respective characteristics, and respective address ranges.

9. The system of claim 1 , wherein the orchestrator is configured to provide the application's requirements to the resource manager for assembling the composite computing node.

10. The system of claim 1 , wherein the resource manager is configured to associate a memory or storage hardware resource with one of the memory or storage tiers only if its associated hardware resource characteristics fall within a predefined range for the respective tier.

11. The system of claim 10 , wherein the resource manager is configured to associate a memory or storage hardware resource with one of the memory or storage tiers only if at least one of its associated processing performance, power consumption, bandwidth, latency, capacity, underlying technology, or cost falls within a predefined range for the respective tier.

12. The system of claim 1 , wherein a first one of the of memory or storage tiers is associated with local memory or storage hardware resources of a compute sled and at least a second one of the of memory or storage tiers is associated with dynamically configured disaggregated computing memory or storage hardware resources.

13. The system of claim 1 , wherein the composite computing node is configured to redirect storage or memory requests targeting one or more of the determined memory or storage tiers to one or more respective memory or storage hardware resources being associated with a respective memory or storage tier.

14. A data center, comprising:

at least one compute sled including a set of local computing hardware resources;

a computing fabric including a plurality of different disaggregated memory and storage hardware resources having respective hardware characteristics;

an orchestrator configured to schedule an application having application specific hardware requirements; and

a data center resource management hardware component having access to the respective hardware characteristics of the different disaggregated memory and storage hardware resources and configured to assemble a composite computing node by selecting one or more disaggregated memory and storage hardware resources with respective hardware characteristics meeting the application's hardware requirements and connecting the compute sled to the selected disaggregated memory and storage hardware resources through the computing fabric,

wherein the orchestrator is configured to determine, according to the application specific hardware requirements, a number of memory and/or storage tiers of the composite computing node, wherein different memory and/or storage tiers are associated with respective different hardware characteristics, and wherein the data center resource management hardware component is configured to associate one or more memory or storage hardware resources meeting the respective hardware characteristic with a respective memory or storage tier and map a respective address range to a respective memory or storage tier.

15. The data center of claim 14 , wherein the data center resource management hardware component includes storage, which is accessible by the orchestrator, for storing metadata indicative of the respective hardware characteristics associated with the plurality of different disaggregated memory and storage hardware resources available in the data center.

16. The data center of claim 14 , wherein the compute sled includes a set of control registers and wherein the data center resource management hardware component is configured to specify the selected disaggregated memory and storage hardware resources by writing to the control registers.

17. A method comprising:

composing a computing system based on a plurality of different disaggregated computing hardware resources having respective hardware characteristics, composing the computing system to include:

storing information on the respective hardware characteristics of the different disaggregated computing hardware resources;

assembling a composite computing node by selecting, based on the stored information, one or more disaggregated computing hardware resources with respective hardware characteristics meeting requirements of an application to be executed on the composite computing node;

determining, according to the requirements of the application, a number of memory and/or storage tiers of the composite computing node, wherein different memory and/or storage tiers are associated with respective different hardware resource characteristics; and

associating one or more memory or storage hardware resources meeting the respective hardware resource characteristic with a respective memory or storage tier and mapping a respective address range to a respective memory or storage tier.

18. The method of claim 17 , wherein respective hardware characteristics comprise processing performance, power, bandwidth, latency, capacity, underlying technology, or cost.

19. The method of claim 17 , wherein the different distributed computing hardware resources comprise different types of computer memory or storage.

20. The method of claim 17 , wherein assembling the composite computing node further comprises

connecting a compute sled with a set of local computing hardware resources to the selected disaggregated computing hardware resources through the, computing fabric.

21. The method of claim 20 , wherein the local computing hardware resources of the compute sled comprise one or more local CPUs and one or more local memory modules.

22. The method of claim 20 , wherein assembling the composite computing node further comprises

writing the respective hardware characteristics of the selected disaggregated computing hardware to one or more control registers of the compute sled.

23. The method of claim 22 , wherein the one or more control registers are indicative of at least one of the groups of a number of memory or storage tiers, respective characteristics, and respective address ranges.

24. The method of claim 17 , further comprising

providing the application's requirements from an orchestrator for scheduling the application to the resource manager for assembling the composite computing node.