Electronic system with learning mechanism and method of operation thereof

Abstract

An electronic system includes: a compute device interface configured to receive system information; a compute device control unit, coupled to the compute device interface, configured to implement: a preprocessing block for partitioning initial data, based on the system information, into first partial data to be processed by a system device and second partial data to be processed by a compute device, and a learning block for processing the second partial data as part of a distributed machine learning processes.

What is claimed is: 
     
         1 . An electronic system comprising:
 a compute device interface configured to receive system information;   a compute device control unit, coupled to the compute device interface, configured to implement:
 a preprocessing block for partitioning initial data, based on the system information, into first partial data to be processed by a system device and second partial data to be processed by a compute device, and 
 a learning block for processing the second partial data as part of a distributed machine learning processes. 
   
     
     
         2 . The system as claimed in  claim 1  wherein the compute device control unit is configured to implement an adaptive preprocessing block. 
     
     
         3 . The system as claimed in  claim 1  further comprising a compute device storage unit, coupled to the compute device control unit, configured to store the initial data as a part of a decentralized storage system of the initial data. 
     
     
         4 . The system as claimed in  claim 1  wherein the compute device control unit is configured to implement a scheduler block, including a per-node scheduler of the compute device for hierarchical coordination of a cluster of further compute devices. 
     
     
         5 . The system as claimed in  claim 1  wherein the compute device control unit is configured to implement the preprocessing block for partitioning the initial data based on a selection ratio. 
     
     
         6 . The system as claimed in  claim 1  wherein the system information includes system utilization, system bandwidth, input/output utilization, or a combination thereof for the system device. 
     
     
         7 . The system as claimed in  claim 1  wherein the system device includes the compute device as a component of the system device for providing a nested distributed machine learning process between the system device and the compute device. 
     
     
         8 . A method of operation of an electronic system comprising:
 receiving system information with a compute device interface;   partitioning initial data, with a compute device control unit, based on the system information, into first partial data to be processed by a system device and second partial data to be processed by a compute device, and   processing the second partial data as part of a distributing machine learning process.   
     
     
         9 . The method as claimed in  claim 8  wherein partitioning the initial data includes partitioning the initial data with an adaptive preprocessing block. 
     
     
         10 . The method as claimed in  claim 8  further comprising storing the initial data on a compute device storage unit, coupled to the compute device control unit, as a part of a decentralized storage system of the initial data. 
     
     
         11 . The method as claimed in  claim 8  further comprising providing hierarchical coordination of a cluster of further compute devices with a scheduler block, including a per-node scheduler of the compute device. 
     
     
         12 . The method as claimed in  claim 8  wherein partitioning the initial data includes partitioning based on a selection ratio. 
     
     
         13 . The method as claimed in  claim 8  wherein receiving the system information includes receiving the system information including system utilization, system bandwidth, input/output utilization, or a combination thereof for the system device. 
     
     
         14 . The method as claimed in  claim 8  wherein processing the second partial data includes processing the second partial as part of a nested distributed machine learning process between the system device and the compute device, wherein the system device includes the compute device as a component of the system device. 
     
     
         15 . A non-transitory computer readable medium including stored thereon instructions to be executed by a control unit comprising:
 receiving system information with a compute device interface;   partitioning initial data, with a compute device control unit, based on the system information, into first partial data to be processed by a system device and second partial data to be processed by a compute device, and   processing the second partial data as part of a distributing machine learning process.   
     
     
         16 . The medium as claimed in  claim 15  wherein partitioning the initial data includes partitioning the initial data with an adaptive preprocessing block. 
     
     
         17 . The medium as claimed in  claim 15  further comprising storing the initial data on a compute device storage unit, coupled to the compute device control unit, as a part of a decentralized storage system of the initial data. 
     
     
         18 . The medium as claimed in  claim 15  further comprising providing hierarchical coordination of a cluster of further compute devices with a scheduler block, including a per-node scheduler of the compute device. 
     
     
         19 . The medium as claimed in  claim 15  wherein receiving the system information includes receiving the system information including system utilization, system bandwidth, input/output utilization, or a combination thereof for the system device. 
     
     
         20 . The medium as claimed in  claim 15  wherein processing the second partial data includes processing the second partial as part of a nested distributed machine learning process between the system device and the compute device, wherein the system device includes the compute device as a component of the system device.