Profiling of software and circuit designs utilizing data operation analyses

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.

Description

CROSS REFERENCE TO A RELATED APPLICATION[0001]This application is related to Paul L. Master et al., U.S. patent application Ser. No. 09 / 815,122, entitled “Adaptive Integrated Circuitry With Heterogeneous And Reconfigurable Matrices Of Diverse And Adaptive Computational Units Having Fixed, Application Specific Computational Elements”, filed Mar. 22, 2001, commonly assigned to QuickSilver Technology, Inc., and incorporated by reference herein, with priority claimed for all commonly disclosed subject matter (the “related application”).FIELD OF THE INVENTION[0002]The present invention relates in general to profiling of software and circuit designs for performance analyses and, more particularly, to profiling both software and reconfigurable and adaptive circuit designs utilizing data parameters, including data dynamics and other data operational statistics.BACKGROUND OF THE INVENTION[0003]Software or other computing programs, such as programs expressed in C and C++ code, have been profi...

AI Technical Summary

Problems solved by technology

Current profiling techniques are generally statistical or intrusive.

Other existing profiling techniques are typically intrusive.

Both statistical profiling and intrusive profiling have significant limitations.

For example, depending upon the granularity or degree to which code has been inserted or interrupts generated, both methodologies may typically miss or overlook code features between such points of intrusion or interrupt.

In addition, measures of power and performance based upon such current statistical or intrusive profiling may be significantly inaccurate.

In addition, such power and performance measures also depends significantly upon program dynamics, such that statistical or intrusive profiling often provides inaccurate results compared to actual performance of the program.

As a consequence, because current profiling techniques do not account for data issues and concerns, they tend to be significantly inaccurate.

Existing profiling tools do not address the impact of each of these variables.

Images