Selecting a number of processing resources to run an application effectively while saving power

1. A method, in a data processing system, for selecting a number of processors to run an application in order to save power, the method comprising:

sampling a number of heavily used parallel code segments from the application;

executing each of the heavily used parallel code segments using two or more of a plurality of processing resource combinations;

scoring each of the heavily used parallel code segments with a performance value, wherein the performance value indicates a performance of each heavily used parallel code segment using each of the two or more processing resource combinations;

selecting one of the two or more processing resource combinations based on an associated performance value and a number of processing resources used to execute the heavily used parallel code segment; and

executing the application using the selected processing resource combination.

2. The method of claim 1 , wherein the performance of each heavily used parallel code segment is a benchmark measurement that rates the speed at which the processing resource combination executes the heavily used parallel code segment.

3. The method of claim 1 , wherein the associated performance value is within a predetermined performance level and wherein the predetermined performance level is within a predetermined percentage of a highest performance value of each heavily used parallel code segment.

4. The method of claim 1 , wherein the plurality of processing resource combinations comprises at least one of two or more physical processors, two or more logical processors, or a combination of physical processors and logical processors.

5. The method of claim 4 , wherein each processor in the two or more physical processors comprises a plurality of cores.

6. The method of claim 5 , wherein the plurality of cores are homogeneous.

7. The method of claim 5 , wherein the plurality of cores are heterogeneous.

8. The method of claim 1 , wherein sampling the number of heavily used parallel code segments from the application is performed prior to the application being executed by an operating system.

9. The method of claim 1 , wherein the plurality of processing resource combinations are predetermined by a user.

10. A computer program product comprising a computer recordable medium having a computer readable program recorded thereon, wherein the computer readable program, when executed on a computing device, causes the computing device to:

sample a number of heavily used parallel code segments from the application;

execute each of the heavily used parallel code segments using two or more of a plurality of processing resource combinations;

score each of the heavily used parallel code segments with a performance value, wherein the performance value indicates a performance of each heavily used parallel code segment using each of the two or more processing resource combinations;

select one of the two or more processing resource combinations based on an associated performance value and a number of processing resources used to execute the heavily used parallel code segment; and

execute the application using the selected processing resource combination.

11. The computer program product of claim 10 , wherein the performance of each heavily used parallel code segment is a benchmark measurement that rates the speed at which the processing resource combination executes the heavily used parallel code segment.

12. The computer program product of claim 10 , wherein the associated performance value is within a predetermined performance level and wherein the predetermined performance level is within a predetermined percentage of a highest performance value of each heavily used parallel code segment.

13. The computer program product of claim 10 , wherein the computer readable program to sample the number of heavily used parallel code segments from the application is performed prior to the application being executed by an operating system.

14. An, apparatus, comprising:

a processor; and

a memory coupled to the processor, wherein the memory comprises instructions which, when executed by the processor, cause the processor to:

sample a number of heavily used parallel code segments from the application;

execute each of the heavily used parallel code segments using two or more of a plurality of processing resource combinations;

score each of the heavily used parallel code segments with a performance value, wherein the performance value indicates a performance of each heavily used parallel code segment using each of the two or more processing resource combinations;

select one of the two or more processing resource combinations based on an associated performance value and a number of processing resources used to execute the heavily used parallel code segment; and

execute the application using the selected processing resource combination.

15. The apparatus of claim 14 , wherein the performance of each heavily used parallel code segment is a benchmark measurement that rates the speed at which the processing resource combination executes the heavily used parallel code segment.

16. The apparatus of claim. 14 , wherein the associated performance value is within a predetermined performance level and wherein the predetermined performance level is within a predetermined percentage of a highest performance value of each heavily used parallel code segment.

17. The apparatus of claim 14 , wherein the instructions to sample the number of heavily used parallel code segments from the application are performed prior to the application being executed by an operating system.