Window: Suitability Report
The upper-left area shows the
Maximum Program Gain for All Sitesin the program. Your overall goal of adding parallelism is to increase the
Maximum Program Gain for All Sitesso the parallel program will execute as fast as possible. The measured serial execution runtime, predicted parallel runtime, and any measured are displayed below
Maximum Program Gain for All Sites. Use the predicted Suitability gain values to help you make informed decisions about where to add parallelism.
If the Suitability tool detects any annotation-related errors, they appear at the top of the
Suitability Reportwindow. If you see this type of error, the displayed Suitability data may not be reliable. Annotation-related errors may be caused when the correct sequence of annotations do not occur because of missing annotations, when unexpected execution paths occur, or if Suitability data collection was paused while the target was executing.
Use the upper-right row of to model performance. Choose a hardware configuration and threading model (parallel framework) values from the drop-down lists. If you select a
Intel® Xeon Phi™processors, an additional value for total
Below this row is a grid of data that shows the estimated performance of each parallel site detected during program execution. The
Site Labelshows the argument to the site annotation. Examine the predicted
Impact to Program Gain(higher values are better) to estimate how much each site contributes to the
Maximum Program Gain for All Sitesfor all sites (described above). To expand the data under
Combined Site Metricsor
Site Instance Metrics, click the icon to the right of that heading; to collapse data, click to the right of that heading.
To view source code for a selected parallel site, click its row to display the
To show or hide the side command toolbar, click the or icon.
Scalability of Maximum Site Gaingraph summarizes performance for the selected site. The number of CPU processors or total number of coprocessor threads appears on the horizontal X axis and the target's predicted performance gain appears on the Y axis. To change the default
CPU Countand the
Maximum CPU Count, set the Options value.
If you choose a
CPU, to view detailed characteristics of the selected site as well as its tasks and locks, click the
Loop Iterations (Tasks) Modeling(or
Tasks Modeling) to experiment with different loop structures, iteration counts, and instance durations that might improve the predicted parallel performance.
For example, you might want to see the impact of modifying your nested change loop structure, modify the loop body code, or change number of iterations.
If the task annotations indicate likely task parallelismhttps://software.intel.com/content/www/us/en/develop/documentation/advisor-user-guide/top/optimize-cpu-usage/threading-perspective/annotate-code-for-deeper-analysis/annotate-code-to-model-parallelism/task-patterns/data-and-task-parallelism.html, the title will appear as
Task Modeling(instead of
Loop Iterations (Task) Modelingfor data parallelism).
Runtime Modelingto learn which parallel overhead categories might have an impact on parallel overhead. If you agree to address a category later by using the chosen parallel framework's capabilities or by tuning the parallel code after you have implemented parallelism, check that category.
If the chosen
Intel Xeon Phior
Offload to Intel Xeon Phi, additional options appear below the
Runtime Modelingarea. To expand this area, click the down arrow to the right of
Intel Xeon Phi Advanced Modeling.
Below the graph is a list of issues that might be preventing better
predictedperformance gains as well as a summary of serial and predicted parallel time. To expand a line, click the down arrow to the right of the item's name. Most issues are related to the
Runtime Modeling. Later, you can use other Analyzer tools like to measure
actualperformance of your parallel program.