Power efficient machine learning in cloud-backed mobile systems
Various embodiments are provided for load balancing of machine learning operations in a computing environment by a processor. One or more machine learning operations performing inference or training operations may by dynamically balanced between one or more edge computing devices in a wireless communication network and a cloud computing system for increasing performance of a selected metric.
1. A method, by a processor, for load balancing of machine learning operations in a computing environment comprising:
dynamically balancing one or more machine learning operations between one or more edge computing devices in a wireless communication network and a cloud computing system for increasing performance of a selected metric, wherein a variable of the one or more machine learning operations is used to determine which of the one or more edge computing devices and the cloud computing system executes the one or more machine learning operations based on an execution time and a predefined threshold, and wherein the variable having an undefined value causes the one or more machine learning operations to execute on both the one or more edge computing devices and the cloud computing system.
2. The method of claim 1 , further including selecting as the selected metric a power metric, temperature metric, a performance metric, data throughput metric, or a combination thereof.
3. The method of claim 1 , further including performing an inference operation or training operation by the one or more machine learning operations.
4. The method of claim 1 , further including determining whether the one or more machine learning operations are executing on the one or more edge computing devices, the cloud computing system, or a combination thereof according to the variable.
5. The method of claim 4 , further including setting the variable as the edge computing device, the cloud computing system, or a combination thereof for indicating one or more current platforms performing an inference operation.
6. The method of claim 4 , further including setting the variable as the edge computing device, the cloud computing system, or a combination thereof according to one or more rules, conditions, or metrics.
7. The method of claim 1 , further including simultaneously performing one or more similar or different inference operations on both the one or more edge computing devices and the cloud computing system.
8. A system for load balancing of machine learning operations in a computing environment, comprising:
one or more computers with executable instructions that when executed cause the system to:
dynamically balance one or more machine learning operations between one or more edge computing devices in a wireless communication network and a cloud computing system for increasing performance of a selected metric, wherein a variable of the one or more machine learning operations is used to determine which of the one or more edge computing devices and the cloud computing system executes the one or more machine learning operations based on an execution time and a predefined threshold, and wherein the variable having an undefined value causes the one or more machine learning operations to execute on both the one or more edge computing devices and the cloud computing system.
9. The system of claim 8 , wherein the executable instructions further select as the selected metric a power metric, temperature metric, a performance metric, data throughput metric, or a combination thereof.
10. The system of claim 8 , wherein the executable instructions further perform an inference operation or training operation by the one or more machine learning operations.
11. The system of claim 8 , wherein the executable instructions further determine whether the one or more machine learning operations are executing on the one or more edge computing devices, the cloud computing system, or a combination thereof according to the variable.
12. The system of claim 11 , wherein the executable instructions further set the variable as the edge computing device, the cloud computing system, or a combination thereof for indicating one or more current platforms performing an inference operation.
13. The system of claim 8 , wherein the executable instructions further set the variable as the edge computing device, the cloud computing system, or a combination thereof according to one or more rules, conditions, or metrics.
14. The system of claim 8 , wherein the executable instructions further simultaneously perform one or more similar or different inference operations on both the one or more edge computing devices and the cloud computing system.
15. A computer program product for load balancing of machine learning operations in a computing environment by a processor, the computer program product comprising a non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising:
an executable portion that dynamically balances one or more machine learning operations between one or more edge computing devices in a wireless communication network and a cloud computing system for increasing performance of a selected metric, wherein a variable of the one or more machine learning operations is used to determine which of the one or more edge computing devices and the cloud computing system executes the one or more machine learning operations based on an execution time and a predefined threshold, and wherein the variable having an undefined value causes the one or more machine learning operations to execute on both the one or more edge computing devices and the cloud computing system.
16. The computer program product of claim 15 , further including an executable portion that selects as the selected metric a power metric, temperature metric, a performance metric, data throughput metric, or a combination thereof.
17. The computer program product of claim 15 , further including an executable portion that performs an inference operation or training operation by the one or more machine learning operations.
18. The computer program product of claim 15 , further including an executable portion that determines whether the one or more machine learning operations are executing on the one or more edge computing devices, the cloud computing system, or a combination thereof according to the variable.
19. The computer program product of claim 18 , further including an executable portion that:
set the variable as the edge computing device, the cloud computing system, or a combination thereof for indicating one or more current platforms performing an inference operation; or
set the variable as the edge computing device, the cloud computing system, or a combination thereof according to one or more rules, conditions, or metrics.
20. The computer program product of claim 15 , further including an executable portion that simultaneously performs one or more similar or different inference operations on both the one or more edge computing devices and the cloud computing system.