Vectorization Advisor - Analysis Summary

Purpose and Usage | Controls | Report Regions

Purpose and Usage

After running Intel® Advisor analyses (Survey, Trip Counts, FLOPS, Memory Access Patterns, Dependencies, and Suitability), consider reviewing a results summary that includes the most important information about your code. Click the Summary tab after running an analysis to view results.

Controls

Use This

To Do This

and toggle buttons

Hide or show information (click).

Site location link

See corresponding line in the report grid (click).

Source location link

See corresponding line in the source file, if available (click).

Recommendation link

See corresponding recommendation in the report window (click).

and buttons

Hide or show loop metric details for MKL application.

Report Regions

Program metrics

The Program metrics section comprises the main performance metrics of your program, such as execution time statistics, vector instruction set, and number of CPU threads utilized. The section also covers performance characteristics such as CPU time and loop execution information. You can find the vectorization efficiency data in this section as well.

Note

The vectorization efficiency data is available only for vectorized loops in modules compiled with an Intel® compiler version 16 or higher.

If your application uses Intel® Math Kernel Library (Intel® MKL), you will see the MKL detail button in the Performance characteristics section, which toggles two additional columns: the User column, which reports time spent in your code and corresponding compute metrics, and the MKL column, which reports time spent in the MKL code and corresponding compute metrics.

If you run the Trip Counts and FLOP or the Roofline analysis, you will see additional information reported: GFLOPS and GINTOPS data and the OP/S and Bandwidth sub-section. The sub-section reports GFLOPS and GINTOPS usage and cache bandwidth metrics compared to hardware peak. In the Utilization column, you can select one of two benchmarks: single-core or multicore.

Per program recommendations

The Per program recommendations section summarizes suggested changes for your program that you might want to apply to achieve better performance.

Top time-consuming loops

The Top time-consuming loops section reports top five loops sorted by self time. The section summarizes the loop names, loop source locations, and execution time statistics. If you run the Trip Counts and FLOP or the Roofline analysis, the section also reports the Trip Counts data.

Refinement analysis data

Note

The information in the Refinement analysis data section is available only after you run the Memory Access Patterns or Dependencies analysis.

The Refinement analysis data provides details about found dependencies and memory access patterns. The Dependencies column summarizes the predicted data sharing problems collected by the Dependencies tool. To display the Dependencies Report window at the corresponding parallel site location, click a function link in the Site Location column.

The Strides Distribution column reports the memory access stride distribution within a loop in the ratio format in %: unit strides, constant strides, and variable strides.

Recommendations

The Recommendations section summarizes suggested changes with high confidence level for first five loops in the code that you might want to apply to achieve better performance. Click a recommendation link to access the recommendations texts.

Collection details

The Collection details section shows execution statistics for each of the collectors, as well as the Collection Log, Application Output, and Collection Command Line links that lead to the corresponding report logs, command line and output details.

Note

Application Output is available if you set output destination to Application Output window. To do this, go to File > Options > General > Application Output Destination and choose Application Output window.

Platform information

The Platform Information section reports the system information including software and hardware summary of your system.

See Also

For more complete information about compiler optimizations, see our Optimization Notice.