Techniques for distributed processing task portion assignment

Abstract

Various embodiments are generally directed to techniques for assigning portions of a task among individual cores of one or more processor components of each processing device of a distributed processing system. An apparatus to assign processor component cores to perform task portions includes a processor component; an interface to couple the processor component to a network to receive data that indicates available cores of base and subsystem processor components of processing devices of a distributed processing system, the subsystem processor components made accessible on the network through the base processor components; and a core selection component for execution by the processor component to select cores from among the available cores to execute instances of task portion routines of a task based on a selected balance point between compute time and power consumption needed to execute the instances of the task portion routines. Other embodiments are described and claimed.

1 . An apparatus to assign processor component cores to perform task portions comprising:
 a processor component;   an interface to couple the processor component to a network to receive data that indicates available cores of base and subsystem processor components of processing devices of a distributed processing system, the subsystem processor components made accessible on the network through the base processor components; and   a core selection component for execution by the processor component to select cores from among the available cores to execute instances of task portion routines of a task based on a selected balance between compute time and power consumption needed to execute the instances of the task portion routines.   
     
     
         2 . The apparatus of  claim 1 , comprising a resource component for execution by the processor component to receive resource data that indicates a quantity of cores present in at least one base processor component and a quantity of cores present in at least one subsystem processor component of a processing device of the distributed processing system. 
     
     
         3 . The apparatus of  claim 2 , the resource data comprising an indication of an amount of storage associated with the at least one base processor component and an amount of storage associated with the at least one subsystem processor component. 
     
     
         4 . The apparatus of  claim 2 , comprising a status component for execution by the processor component to receive a status data that indicates a subset of the quantity of cores present in the at least one base processor component that are available for assignment of a task portion and indicates a subset of the quantity of cores present in the at least one subsystem processor component that are available for assignment of a task portion. 
     
     
         5 . The apparatus of  claim 1 , comprising a status component for execution by the processor component to receive a first status data that indicates a current status of execution of an instance of a task portion routine of the task by one of the selected cores. 
     
     
         6 . The apparatus of  claim 5 , the status component to receive a query from a submission device for a current status of performance of the task and to transmit a second status data that indicates the current status of performance of the task based on a current status of execution of the instances of the task portion routines by the selected cores. 
     
     
         7 . The apparatus of  claim 1 , the core selection component to determine relative periods of time required to perform each task portion of the task, and to select a type of core from among types of cores included in the available cores to perform each task portion based on the relative periods of time. 
     
     
         8 . The apparatus of  claim 7 , the core selection component to select the selected cores from among the available cores to assign to perform instances of a task portion that requires a relatively long period of time to perform in response to the selected balance comprising favoring reducing the electric power consumed over reducing the time required. 
     
     
         9 . The apparatus of  claim 1 , the core selection component to select the selected cores from among the available cores to increase a degree of locality among the selected cores to minimize inter-process communications between ones of the processing devices in response to the selected balance comprising favoring reducing the time required over reducing the electric power consumed. 
     
     
         10 . The apparatus of  claim 1 , the core selection component to transmit indications of the selection of the selected cores via the interface to the selected cores. 
     
     
         11 . An apparatus to execute task portion routines comprising:
 a base processor component;   an interface to couple the base processor component to a network;   a subsystem processor component;   a bus to communicatively couple the subsystem processor component to the base processor component; and   a routing component for execution by the base processor component to receive a packet from the network, to determine the Internet protocol (IP) address to which the packet is directed, and to route the packet to the subsystem processor component via the bus based on the IP address.   
     
     
         12 . The apparatus of  claim 11 , comprising:
 a first resource component for execution by the base processor component to transmit a first resource data that indicates a quantity of cores present in the base processor in a first packet on the network, the first packet indicating a first IP address associated with the base processor as its source; and   a second resource component for execution by the subsystem processor component to transmit a second resource data that indicates a quantity of cores present in the subsystem processor in a second packet to the routing component via the bus, the routing component to transmit the second packet on the network, the second packet indicating a second IP address associated with the subsystem processor as its source.   
     
     
         13 . The apparatus of  claim 11 , comprising:
 a first communications component for execution by the base processor component to present the base processor component on the network at a first IP address to enable the base processor component to receive packets directed to the first IP address, and to accept the packet at the base processor component based on the IP address to which the packet is directed comprising the first IP address; and   a second communications component for execution by the subsystem processor component to accept the packet at the subsystem processor component based on the IP address to which the packet is directed comprising a second IP address, the routing component to present the subsystem processor component on the network at the second IP address to enable the subsystem processor component to receive packets direct to the second IP address.   
     
     
         14 . The apparatus of  claim 13 , comprising:
 a first execution component for execution by the base processor component to execute a first task portion routine received from the network in at least one packet directed to the first IP address; and   a second execution component for execution by the subsystem processor component to execute a second task portion routine received from the network in at least one packet directed to the second IP address.   
     
     
         15 . The apparatus of  claim 14 , the first execution component to transmit a first status data indicating status of execution of the first task portion to a scheduling server via the network in a packet indicating the first IP address as a source. 
     
     
         16 . The apparatus of  claim 15 , the second execution component to transmit a second status data indicating status of execution of the second task portion to the routing component, and the routing component to transmit the second status data indicating status of execution of the second task portion to the scheduling server via the network in a packet indicating the second IP address as a source. 
     
     
         17 . A computing-implemented method for assigning processor component cores to perform task portions comprising:
 receiving status data indicating available cores of base and subsystem processor components of processing devices of a distributed processing system, the subsystem processor components made accessible on the network through the base processor components;   selecting cores from among the available cores to execute instances of task portion routines of a task based on a selected trade-off between reducing time required and reducing electric power consumed to execute the instances of the task portion routines; and   transmitting indications of the selection of the selected cores via the interface to the selected cores.   
     
     
         18 . The computer-implemented method of  claim 17 , comprising receiving resource data indicating resources associated with each base processor component and each subsystem processor component of each processing device of the distributed processing system. 
     
     
         19 . The computer-implemented method of  claim 18 , the resources associated with each base processor component and each subsystem processor component comprising one of a quantity of cores present in at least one base processor component or in at least one subsystem processor component, a size of a cache of a core of a base processor component or a subsystem processor component, or an amount of storage associated with a base processor component or a subsystem processor component. 
     
     
         20 . The computer-implemented method of  claim 19 , comprising receiving a status data that indicates a subset of the quantity of cores present in the at least one base processor component or the at least one subsystem processor component that are available for assignment of a task portion. 
     
     
         21 . The computer-implemented method of  claim 17 , comprising:
 receiving a first status data that indicates a current status of execution of an instance of a task portion routine of the task by one of the selected cores;   receiving a query from a submission device for a current status of performance of the task; and   transmitting via the interface a second status data that indicates the current status of performance of the task based on a current status of execution of the instances of the task portion routines by the selected cores.   
     
     
         22 . The computer-implemented method of  claim 17 , comprising:
 determining relative periods of time required to perform each task portion of the task; and   selecting a type of core from among types of cores included in the available cores to perform each task portion based on the relative periods of time.   
     
     
         23 . The computer-implemented method of  claim 22 , comprising selecting the selected cores from among the available cores to assign to perform instances of a task portion that requires a relatively long period of time to perform in response to the selected trade-off comprising favoring reducing the electric power consumed over reducing the time required. 
     
     
         24 . The computer-implemented method of  claim 17 , comprising selecting the selected cores from among the available cores to increase a degree of locality among the selected cores to minimize inter-process communications between ones of the processing devices in response to the selected trade-off comprising favoring reducing the time required over reducing the electric power consumed. 
     
     
         25 . At least one machine-readable storage medium comprising instructions that when executed by a computing device, cause the computing device to perform the method of any of  claims 17 - 24 .