Resource modeling and scheduling for extensible computing platforms

摘要

Energy management modeling and scheduling techniques are described for reducing the power consumed to execute an application on a multi-processor computing platform within a certain time period. In one embodiment, a sophisticated resource model which accounts for discrete operating modes for computing components/resources on a computing platform and transition costs for transitioning between each of the discrete modes is described. This resource model provides information for a specific heterogeneous multi-processor computing platform and an application being implemented on the platform in a form that can be processed by a selection module, typically utilizing an integer linear programming (ILP) solver or algorithm, to select a task schedule and operating configuration(s) for executing the application within a given time.

主权项

1. A method comprising: under control of one or more processors configured with executable instructions, configuring a computer to complete execution of an application within a given time period and using a total amount of power within a power consumption constraint, wherein the configuring comprises:
representing the application as one or more executable tasks; ascertaining operating modes of a plurality of computing resources on the computer, the operating modes comprising at least a standby mode and an active mode, the operating modes being associated with mode parameters, the mode parameters comprising transition cost parameters associated with transitioning between the operating modes, the transition cost parameters comprising a time delay parameter and a transition power consumption parameter, the transition power consumption parameter being associated with a transition from a standby mode to an active mode of a computing resource of the plurality of computing resources; and selecting, based at least in part on one or more of the mode parameters: one or more computing resources from the plurality of computing resources; and one or more of the operating modes for each of the selected one or more computing resources to execute each of the one or more executable tasks, wherein the selecting comprises: mapping the one or more executable tasks with one or more of the operating modes to create a plurality of mappings; and identifying a combination of the plurality of mappings associated with a lowest overall power consumption for executing the application within the given time period.