BEST PERFORMANCE DELIVERY IN HETEROGENEOUS COMPUTING UNIT ENVIRONMENT

• US 20170277550A1
• Filed: 03/22/2016
• Published: 09/28/2017
• Est. Priority Date: 03/22/2016
• Status: Active Grant

• Current Assignee: Lenovo Enterprise Solutions (Singapore) Pte., Ltd. (Lenovo Group Ltd.)
• Original Assignee: Lenovo Enterprise Solutions (Singapore) Pte., Ltd. (Lenovo Group Ltd.)
• Inventors: ZHANG, JIANBANG, NICHOLSON, JOHN W., VANOVER, MICHAEL T.
• Granted Patent: US 10,884,761 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 11/3024 where the computing system component is a central processing unit [CPU]
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 2209/501 Performance criteria
• CPC Class Codes: G06F 9/44505 Configuring for program initiating, e.g. using registry, configuration files
• CPC Class Codes: G06F 9/4484 Executing subprograms
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals

Abstract

An apparatus for selecting an efficient processor includes a comparison module that compares performance characteristics of a plurality of processors available for execution of a function, where each performance characteristic varies as a function of function size. The apparatus includes a selection module that selects, based on a size of the function, a processor from the plurality of processors with a best performance for execution of the function, and an execution module that executes the function on the selected processor.

Claims

1. An apparatus comprising:

a comparison module that compares performance characteristics of a plurality of processors available for execution of a function, wherein each performance characteristic varies as a function of function size;

a selection module that selects, based on a size of the function, a processor from the plurality of processors with a best performance for execution of the function; and

an execution module that executes the function on the selected processor.

2. The apparatus of claim 1, further comprising a performance module that determines the performance characteristic of each of the plurality of processors.

3. The apparatus of claim 2, wherein the performance module further comprises:

a performance calculation module that calculates performance of an executed function executed on a processor; and

a correlation module that correlates a size of each executed function to performance during execution of the function to determine a performance characteristic of the processor.

4. The apparatus of claim 2, further comprising a context module that tracks a context for each function executed on a processor and wherein the comparison module compares performance characteristics of the plurality of processors available for execution of a function with similar contexts.

5. The apparatus of claim 2, wherein the performance module determines the performance characteristic of each of the plurality of processors based on one or more of:

information supplied with the function;

information from a vendor; and

function execution data from previously executed functions.

6. The apparatus of claim 2, wherein the performance module further comprises an update module that updates an initial performance characteristic of a processor based on the function execution data from previously executed functions.

7. The apparatus of claim 1, wherein the selection module selects multiple processors from the plurality of processors with similar best performance characteristics, based on the size of the function, for execution of the function and wherein the execution module executes the function on the multiple selected processors.

8. The apparatus of claim 7, wherein the selection module uses one or more of information from a call stack and a return address to determine a context of a function.

9. The apparatus of claim 1, further comprising a format module that, in response to the selection module selecting a processor for execution of the function, formats the function for execution on the selected processor, wherein the function is formatted for a central-processing unit (“

CPU”

) prior to formatting by the format module.

10. The apparatus of claim 1, wherein the performance characteristics of the plurality of processors comprise one or more of speed of execution, energy efficiency, and cost of execution of a functions of various sizes.

11. The apparatus of claim 1, wherein each of the plurality of processors comprise one or more of a field-programmable gate array (“

FPGA”

), a graphics processing unit (“

GPU”

), an accelerated processing unit (“

APU”

), a central processing unit (“

CPU”

), a digital signal processor (“

DSP”

) and a processor of a computing device different from a computing device executing an application comprising the function.

12. A method comprising:

comparing performance characteristics of a plurality of processors available for execution of a function, wherein each performance characteristic varies as a function of function size;

selecting, based on a size of the function, a processor from the plurality of processors with a best performance for execution of the function; and

executing the function on the selected processor.

13. The method of claim 12, further comprising determining the performance characteristic of each of the plurality of processors.

14. The apparatus of claim 13, wherein determining the performance characteristic of each of the plurality of processors further comprises:

calculating performance of an executed function executed on a processor; and

correlating a size of each executed function to performance during execution of the function to determine a performance characteristic of the processor.

15. The method of claim 14, further comprising tracking a context for each function executed on a processor and wherein comparing performance characteristics of a plurality of processors available for execution of a function comprises comparing performance characteristics of the plurality of processors available for execution of a function with similar contexts.

16. The method of claim 13, wherein determining the performance characteristic of each of the plurality of processors is based on one or more of:

information supplied with the function;

information from a vendor associated with the processor; and

function execution data from previously executed functions.

17. The method of claim 13, wherein determining the performance characteristic of each of the plurality of processors further comprises updating an initial performance characteristic of a processor based on the function execution data from previously executed functions.

18. A program product comprising a computer readable storage medium that stores code executable by a processor, the executable code comprising code to perform:

comparing performance characteristics of a plurality of processors available for execution of a function, wherein each performance characteristic varies as a function of function size;

selecting, based on a size of the function, a processor from the plurality of processors with a best performance for execution of the function; and

executing the function on the selected processor.

19. The program product of claim 18, the code further comprises determining the performance characteristic of each of the plurality of processors.

20. The program product of claim 19, wherein determining the performance characteristic of each of the plurality of processors further comprises:

calculating performance of an executed function executed on a processor; and

correlating a size of each executed function to performance during execution of the function to determine a performance characteristic of the processor.