Profiling of software and circuit designs utilizing data operation analyses

• US 20040093589A1
• Filed: 11/07/2002
• Published: 05/13/2004
• Est. Priority Date: 11/07/2002
• Status: Active Grant

• Current Assignee: Cornami, Inc.
• Original Assignee: QuickSilver Controls, Inc.
• Inventors: Master, Paul L.
• Granted Patent: US 8,276,135 B2
• Time in Patent Office: Days
• US Class Current: 717/136
• CPC Class Codes: G06F 30/33 Design verification, e.g. functional simulation or model checking

Abstract

The present invention is a method, system, software and data structure for profiling programs, other code, and adaptive computing integrated circuit architectures, using a plurality of data parameters such as data type, input and output data size, data source and destination locations, data pipeline length, locality of reference, distance of data movement, speed of data movement, data access frequency, number of data load/stores, memory usage, and data persistence. The profiler of the invention accepts a data set as input, and profiles a plurality of functions by measuring a plurality of data parameters for each function, during operation of the plurality of functions with the input data set, to form a plurality of measured data parameters. From the plurality of measured data parameters, the profiler generates a plurality of data parameter comparative results corresponding to the plurality of functions and the input data set. Based upon the measured data parameters, portions of the profiled code are selected for embodiment as computational elements in an adaptive computing IC architecture.

Claims

1. A method for profiling operation of a plurality of functions, the method comprising:

providing an input data set;

measuring a plurality of data parameters, for each function of the plurality of functions, during operation of the plurality of functions with the input data set, to form a plurality of measured data parameters; and

from the plurality of measured data parameters, generating a plurality of data parameter comparative results corresponding to the plurality of functions and the input data set.

2. The method of claim 1, wherein the plurality of data parameters comprises at least two of the following data parameters:

data location for static data;

data type;

input data size;

output data size;

data source location;

data destination location;

data pipeline length;

locality of reference;

distance of data movement;

speed of data movement;

data access frequency;

number of data load/stores;

cache usage;

register usage;

memory usage, and data persistence.

3. The method of claim 1, further comprising:

determining a corresponding algorithmic element for each function of the plurality of functions.

4. The method of claim 3, further comprising:

determining a computational element of an adapting computing architecture for the corresponding algorithmic element.

5. The method of claim 4, further comprising:

determining configuration information to direct a configuration of a plurality of computational elements of an adapting computing architecture for performance of the corresponding algorithmic element.

6. The method of claim 3, further comprising:

determining a computational element of an adapting computing architecture for the corresponding algorithmic element for each function of the plurality of functions until all functions of the plurality of functions are capable of embodiment in a plurality of heterogeneous computational elements of an adaptive computing architecture, the plurality of heterogeneous computational elements comprised of a corresponding plurality of fixed circuit architectures.

7. The method of claim 3, further comprising:

determining a computational element of an adapting computing architecture for the corresponding algorithmic element for each function of the plurality of functions until all functions of the plurality of functions are capable of embodiment in a processor and in a plurality of heterogeneous computational elements of an adaptive computing architecture coupled to the processor, the plurality of heterogeneous computational elements comprised of a corresponding plurality of fixed circuit architectures.

8. The method of claim 3, further comprising:

determining an integrated circuit location for the corresponding algorithmic element for each function of the plurality of functions.

9. The method of claim 1, wherein the plurality of functions are embodied as program code.

10. The method of claim 9, wherein the measuring of the plurality of data parameters is with reference to a predetermined integrated circuit architecture.

11. The method of claim 1, wherein the plurality of functions are embodied as a plurality of corresponding computational units which perform the plurality of functions within an adaptive computing circuit architecture.

12. The method of claim 1, wherein the plurality of data parameter comparative results are provided in data operational units.

13. The method of claim 12, wherein the plurality of data parameter comparative results in data operational units is a weighted function of the plurality of measured data parameters.

14. The method of claim 1, further comprising:

using the plurality of measured data parameters, determining power consumption for each function of the plurality of functions.

15. The method of claim 1, wherein the measuring of the plurality of data parameters is performed on a relative and comparative basis.

16. The method of claim 1, wherein the measuring of the plurality of data parameters is performed on an empirical basis.

17. The method of claim 1, further comprising, for each function of the plurality of functions:

determining an elapsed time required for performance of each function;

determining the average memory throughput for performance of each function; and

determining a number and type of instructions of a program code performed by each functions.

18. The method of claim 1, wherein the method is performed by a computer.

19. The method of claim 1, wherein the method is performed by an adaptive computing circuit having a plurality of computational units which perform the plurality of functions.

21. The method of claim 1, further comprising:

providing a data structure containing a designation of the plurality of data parameters, the plurality of functions, the input data set, the plurality of measured data parameters, and the plurality of data parameter comparative results.

22. The method of claim 21, further comprising:

visually displaying the data structure.

20. The method of claim 20, wherein the adaptive computing circuit modifies its configuration on the basis of the plurality of data parameter comparative results.

23. A tangible medium storing machine readable software for profiling operation of a plurality of functions, the tangible medium storing machine readable software comprising:

first software for providing an input data set;

second software for measuring a plurality of data parameters, for each function of the plurality of functions, during operation of the plurality of functions with the input data set, to form a plurality of measured data parameters; and

third software for generating, from the plurality of measured data parameter results, a plurality of data parameter comparative results corresponding to the plurality of functions and the input data set.

24. The tangible medium storing machine readable software of claim 23, wherein the plurality of data parameters comprises at least two of the following data parameters:

data location for static data;

data type;

input data size;

output data size;

data source location;

data destination location;

data pipeline length;

locality of reference;

distance of data movement;

speed of data movement;

data access frequency;

number of data load/stores;

cache usage;

register usage;

memory usage, and data persistence.

25. The tangible medium storing machine readable software of claim 23, further comprising:

fourth software for determining a corresponding algorithmic element for each function of the plurality of functions.

26. The tangible medium storing machine readable software of claim 25, further comprising:

fifth software for determining a computational element of an adapting computing architecture for the corresponding algorithmic element.

27. The tangible medium storing machine readable software of claim 26, further comprising:

sixth software for determining configuration information to direct a configuration of a plurality of computational elements of an adapting computing architecture for performance of the corresponding algorithmic element.

28. The tangible medium storing machine readable software of claim 25, further comprising:

seventh software for determining an integrated circuit location for the corresponding algorithmic element for each function of the plurality of functions.

29. The tangible medium storing machine readable software of claim 23, wherein the plurality of functions are embodied as program code.

30. The tangible medium storing machine readable software of claim 23, wherein the plurality of functions are embodied as a plurality of corresponding computational units which perform the plurality of functions within an adaptive computing circuit architecture.

31. The tangible medium storing machine readable software of claim 23, wherein the plurality of data parameter comparative results are provided in data operational units as a weighted function of the plurality of measured data parameters.

32. The tangible medium storing machine readable software of claim 23, further comprising:

eighth software for determining power consumption, for each function of the plurality of functions, using the plurality of measured data parameters.

33. The tangible medium storing machine readable software of claim 23, further comprising, for each function of the plurality of functions:

ninth software for determining an elapsed time required for performance of each function, determining the average memory throughput for performance of each function, and determining a number and type of instructions of a program code performed by each functions.

34. The tangible medium storing machine readable software of claim 23, further comprising:

tenth software for providing a data structure containing a designation of the plurality of data parameters, a designation of the plurality of functions, a designation of the input data set, the plurality of measured data parameters, and the plurality of data parameter comparative results.

35. The tangible medium storing machine readable software of claim 34, further comprising:

eleventh software for visually displaying the data structure.

36. The tangible medium storing machine readable software of claim 23, provided as a computer data signal embodied in a carrier wave.

37. A system for profiling operation of a plurality of functions, the system comprising:

means for providing an input data set;

means for measuring a plurality of data parameters, for each function of the plurality of functions, during operation of the plurality of functions with the input data set, to form a plurality of measured data parameters; and

means for generating, from the plurality of measured data parameters, a plurality of data parameter comparative results corresponding to the plurality of functions and the input data set.

38. The system of claim 37, wherein the plurality of data parameters comprises at least two of the following data parameters:

data location for static data;

data type;

input data size;

output data size;

data source location;

data destination location;

data pipeline length;

locality of reference;

distance of data movement;

speed of data movement;

data access frequency;

number of data load/stores;

cache usage;

register usage;

memory usage, and data persistence.

39. The system of claim 37, further comprising:

means for determining a corresponding algorithmic element for each function of the plurality of functions.

40. The system of claim 39, further comprising:

means for determining a computational element of an adapting computing architecture for the corresponding algorithmic element.

41. The system of claim 40, further comprising:

means for determining configuration information to direct a configuration of a plurality of computational elements of an adapting computing architecture for performance of the corresponding algorithmic element.

42. The system of claim 39, further comprising:

means for determining a computational element of an adapting computing architecture for the corresponding algorithmic element for each function of the plurality of functions until all functions of the plurality of functions are capable of embodiment in a plurality of heterogeneous computational elements of an adaptive computing architecture, the plurality of heterogeneous computational elements comprised of a corresponding plurality of fixed circuit architectures.

43. The system of claim 39, further comprising:

means for determining a computational element of an adapting computing architecture for the corresponding algorithmic element for each function of the plurality of functions until all functions of the plurality of functions are capable of embodiment in a processor and in a plurality of heterogeneous computational elements of an adaptive computing architecture coupled to the processor, the plurality of heterogeneous computational elements comprised of a corresponding plurality of fixed circuit architectures.

44. The system of claim 39, further comprising:

means for determining an integrated circuit location for the corresponding algorithmic element for each function of the plurality of functions.

45. The system of claim 37, wherein the plurality of functions are embodied as program code.

46. The system of claim 37, wherein the plurality of functions are embodied as a plurality of corresponding computational units which perform the plurality of functions within an adaptive computing circuit architecture.

47. The system of claim 37, wherein the plurality of data parameter comparative results are provided in data operational units.

48. The system of claim 47, wherein the plurality of data parameter comparative results in data operational units is a weighted function of the plurality of measured data parameters.

49. The system of claim 37, further comprising:

means for determining power consumption, for each function of the plurality of functions, using the plurality of measured data parameters.

50. The system of claim 37, further comprising, for each function of the plurality of functions:

means for determining an elapsed time required for performance of each function;

means for determining the average memory throughput for performance of each function; and

means for determining a number and type of instructions of a program code performed by each functions.

51. The system of claim 37, wherein the system is embodied as a computer.

52. The system of claim 37, wherein the system is embodied as an adaptive computing circuit having a plurality of computational units which perform the plurality of functions.

53. The system of claim 52, wherein the adaptive computing circuit modifies its configuration on the basis of the plurality of data parameter comparative results.

54. The system of claim 37, further comprising:

means for providing a data structure containing a designation of the plurality of data parameters, the plurality of functions, the input data set, the plurality of measured data parameters, and the plurality of data parameter comparative results.

55. The system of claim 54, further comprising:

means for visually displaying the data structure.

56. A machine-readable medium having stored thereon a date structure comprising:

a first field containing data representing a plurality of data parameters;

a second field containing data representing a plurality of functions;

a third data field containing data representing a plurality of measured data parameters corresponding to the plurality of functions and an input data set; and

a fourth data field containing a plurality of data parameter comparative results corresponding to the plurality of functions and an input data set.

57. The machine-readable medium of claim 56, wherein the plurality of data parameters comprises at least two of the following data parameters:

data location for static data;

data type;

input data size;

output data size;

data source location;

data destination location;

data pipeline length;

locality of reference;

distance of data movement;

speed of data movement;

data access frequency;

number of data load/stores;

cache usage;

register usage;

memory usage, and data persistence.

58. The machine-readable medium of claim 56, wherein each data parameter comparative result, of the plurality of data parameter comparative results, is formed as a weighted function of the plurality of measured data parameters and is expressed in data operational units.

59. The machine-readable medium of claim 56, wherein the third data field further contains data representing a plurality of algorithmic elements corresponding to the plurality of functions.

60. The machine-readable medium of claim 59, wherein the third data field further contains data representing a plurality of computational elements corresponding to the plurality of algorithmic elements.