ENERGY CAPTURE OF TIME-VARYING ENERGY SOURCES BY VARYING COMPUATIONAL WORKLOAD

• US 20120117411A1
• Filed: 11/04/2011
• Published: 05/10/2012
• Est. Priority Date: 11/09/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Stanley-Marbell, Phillip
• Granted Patent: US 8,856,576 B2
• Time in Patent Office: Days
• US Class Current: 713/340
• CPC Class Codes: G06F 1/324 by lowering clock frequency
• CPC Class Codes: G06F 1/329 by task scheduling
• CPC Class Codes: G06F 1/3296 by lowering the supply or operating voltage
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management
• CPC Class Codes: Y10S 323/906 Solar cell systems

Abstract

A method and system for managing computational workload in a computerized system powered by an energy source. This invention relies first and foremost on adapting computational workload at the computerized system according to a time-varying property of the power supplied by the energy source, such that power generated at the source is optimized. A feedback mechanism is accordingly implemented which changes power available for computation, e.g., to track a more efficient energy generation efficiency point at the source.

Claims

1. A method of managing computational workload in a computerized system powered by an energy source, comprising:

adapting said computational workload at said computerized system according to a time-varying property of the power supplied by said energy source, wherein said power generated at said source is optimized.

2. The method according to claim 1, further comprising:

monitoring said property, wherein adapting said computational workload is carried out according to said property as monitored.

3. The method according to claim 1,

wherein time variations of said property are at least partly predetermined;

and

wherein adapting said computational workload is at least partly carried out according to said time variations of said property as predetermined.

4. The method according to claim 1, wherein adapting said computational workload comprises varying a rate of computation by varying at least one of:

a job scheduling at one or more processors of said computerized system;

a frequency f of execution at one or more processors while not modifying a voltage V_dd at said one or more processors; and

a frequency f of execution at one or more processors and correspondingly adjusting a voltage V_dd at said one or more processors, according to a relation wherein said voltage V_dd is made optimal at said one or more processors as a function of a given frequency.

5. The method according to claim 1, wherein said time-varying property is a function of an optimal voltage V_MPP at an output of said energy source.

6. The method according to claim 5, wherein said optimal voltage V_MPP is determined according to a maximum power point of said energy source.

7. The method according to claim 1, wherein:

said computerized system is powered by said energy source through a charge storage device; and

said adapting further comprises varying said computational workload at said computerized system to deplete said charge storage device according to said time-varying property.

8. The method according to claim 7, further comprising:

obtaining, prior to said adapting, a maximum power point of said energy source;

wherein an optimal terminal voltage at said charge storage device and, in turn, an optimal voltage V_MPP at an output of said energy source are determined; and

wherein said adapting comprises varying said computational workload according to said determined optimal voltage V_MPP at said output of said energy source.

9. The method according to claim 1, wherein said adapting comprises:

comparing an optimal voltage V_MPP with an actual voltage V_s at an output of said energy source; and

varying said computational workload according to a result of said comparing.

10. The method according to claim 9, wherein said varying further comprises at least one of:

increasing a rate of computational workload, wherein said actual voltage V_s is larger than said optimal voltage V_MPP; and

decreasing said rate of computational workload, wherein said actual voltage V_s is smaller than said optimal voltage V_MPP.

11. The method according to claim 1, wherein:

said computerized system is powered by said energy source through a charge storage device and a voltage regulator; and

said adapting comprises varying said computational workload to affect a current drawn by said voltage regulator, to deplete said charge storage device according to said time-varying property of said energy source.

12. The method according to claim 1, wherein said energy source comprises a solar energy source.

13. The method according to claim 12, wherein said solar energy source further comprises a photovoltaic system.

14. The method according to claim 1, further comprising:

receiving at said computerized system, from a remote computerized system, additional data to be computed; and

wherein said adapting comprises using said received additional data to vary said computational workload.

15. A computer program product for managing computational workload in a computerized system powered by an energy source, the computer program product comprising:

a computer readable storage medium having computer readable non-transient program code embodied therein, the computer readable program code comprising;

computer readable program code configured to perform the step of;

adapting said computational workload at said computerized system according to a time-varying property of the power supplied by said energy source, wherein said power generated at said source is optimized.

16. A system for managing computational workload in a computerized system powered by an energy source, comprising:

means for adapting said computational workload at said computerized system according to a time-varying property of the power supplied by said energy source, wherein said power generated at said source is optimized.

17. The system according to claim 16, further comprising a charge storage device, wherein said computerized system is powered by said energy source through said charge storage device.