Predictive computing device power management

• US 8,732,487 B2
• Filed: 06/30/2010
• Issued: 05/20/2014
• Est. Priority Date: 06/30/2010
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Goraczko, Michel, Liu, Jie, Kansal, Aman
• Primary Examiner(s): Rehman, Mohammed
• Assistant Examiner(s): Mazyad, Alyaa T
• Application Number:
  US12/826,692
• Publication Number:
  US 20120005490A1
• Time in Patent Office: 1,420 Days
• Field of Search: 713/300, 713/320, 713/310
• US Class Current: 713/300
• CPC Class Codes: G06F 1/3206 Monitoring of events, devices or parameters that trigger a change in power modality

Abstract

The described implementations relate to predictive computing device energy management. One implementation measures resource usage of a computing device that employs a power policy. This implementation also estimates resource usage of the computing device having at least one different power policy without actually running the at least one different power policy on the computing device.

Claims

1. A method, comprising:

measuring resource usage associated with user activity on a computing device that employs an existing power policy that has first power management settings, wherein the measuring comprises determining a function of performance counters from a performance log of the user activity;

obtaining a power model for the computing device that reflects the measured resource usage associated with the user activity and the existing power policy; and

,

utilizing the power model to estimate the resource usage of the computing device having at least one different power policy that has second power management settings, wherein the estimating is accomplished without running the at least one different power policy on the computing device by computing differences in individual performance counters associated with the user activity between the second power management settings of the at least one different power policy and the first power management settings of the existing power policy to simulate running the at least one different power policy, wherein the measuring and the utilizing are performed by a processor.

2. The method of claim 1, wherein the measuring comprises identifying elements that contribute to the resource usage by the computing device and identifying individual power management settings of individual elements, and wherein the existing power policy comprises the identified individual power management settings.

3. The method of claim 2, wherein power is estimated for multiple different power policies including the at least one different power policy, and further comprising recommending an individual different power policy based at least in part on:

user history, a user defined parameter, or third party power policy information.

4. The method of claim 1, wherein the measuring and the estimating are performed relative to the computing device and further comprising sharing the at least one different power policy with other computing devices.

5. The method of claim 1, wherein the at least one different power policy includes power settings obtained from other similar computer devices, another computing device associated with a same user as the computing device, or further computing devices associated with users that are similarly situated to the user of the computing device.

20. The method of claim 1, wherein the obtaining comprises generating.

6. A system, comprising:

a logging module configured to record a log of user activity and energy use associated with that user activity on a computing device employing a power policy of power management settings;

a power modeling module configured to generate a power model for the computing device based upon the user activity and the associated energy use;

a predictive power management tool configured to automatically recommend new power management settings of a new power policy for the computing device based on the log; and

,

wherein the predictive power management tool is further configured to predict an impact on energy use relating to changing to the new power management settings for the computing device utilizing the log, the power model and differences between the power management settings and the new power management settings; and

a processor configured to execute at least one of the logging module, the power modeling module, or the predictive power management tool.

7. The system of claim 6, wherein the predictive power management tool is configured to obtain power management setting information from other computing devices that are similar to the computing device.

8. The system of claim 6, wherein the predictive power management tool is configured to predict the impact without running the new power management settings on the computing device.

9. The system of claim 6, wherein the predictive power management tool comprises an energy simulation engine configured to predict the impact by estimating potential energy usage under the new power policy by running a simulation based on the log, the power model and differences between the power management settings and the new power management settings.

10. The system of claim 6, wherein the system comprises the computing device or wherein the system comprises the computing device and at least one other computing device and wherein the computing device and the at least one other computing device are associated with a single user.

18. The system of claim 6, wherein the system is included on the computing device.

19. The system of claim 6, wherein the system does not include the computing device.

11. A method, comprising:

obtaining a power model of a computer device, the power model reflecting energy use associated with individual features of the computing device according to present power settings;

tracking resource usage during user activity of the computing device;

remapping resources related to changing to different power settings of the computing device; and

,

estimating power consumption associated with the changed power settings without running the computing device with the changed power settings by utilizing the power model and a difference between the tracked resource usage associated with the individual features at the present power settings compared to the individual features at the changed power settings for the user activity, wherein at least the estimating is performed by a processor.

12. The method of claim 11, wherein the obtaining the power model comprises generating the power model with a software-based mapping from the resource usage to the power consumption.

13. The method of claim 11, wherein the tracking comprises generating a log of the resource usage and further comprising filtering the log and recording only a sub-set of the log.

14. The method of claim 11, wherein the tracking comprises generating a log of the resource usage and further comprising compressing the log and recording the compressed log.

15. The method of claim 11, wherein the present power settings are changed to satisfy a user defined parameter.

16. The method of claim 15, wherein the user defined parameter relates to a desired run time of the computing device.

17. The method of claim 11, wherein responsive to the tracking, further comprising utilizing an increase in the power consumption as an indication to check for a failing component.