User Guide

  • 2020
  • 06/18/2020
  • Public Content
Contents

Suitability Report, Site Row

The Site rows summarize the predicted performance improvement for each parallel site you have annotated in your application. If only one annotated parallel site was executed, only a single site row will be present.
Selecting a Site row allows you to view the characteristics of the selected parallel site and choose its modeling parameters, including the scalability graph and likely causes, modeling of tasks and task iterations, and runtime impact of overhead and contention items. Based on your current annotations, the check boxes under
Runtime Modeling
show what would happen if you agree to fix different types of parallel overhead items when you add parallelism.
View the upper-left corner. Your goal is to make the
Site Gain
as close to the
CPU
count as possible. Improving each site's
Site Gain
will improve the combined
Maximum Program Gain for All Sites
.
If you see an Annotation Label named
_too_deep_
, this indicates that the call behavior of your program results in nesting of site and/or task annotations more deeply than
Intel Advisor
can handle. The deeper occurrences have all been merged into this
_too_deep_
pseudo-site. This can indicate a problem with the placement of the annotation pair. For example, deep recursions or not executing site-end or task-end annotations can result in such nesting.

For Each Site, Decide Whether to Modify or Keep Annotations

Use the
Suitability Report
window to view the predicted parallel performance of each parallel site and its impact on the
Maximum Program Gain for All Sites
. For example, if a site either has a
Site Gain
of less than 1.0 or does not contribute to
Maximum Program Gain for All Sites
, modify or remove its annotations. In contrast, any site that contributes to
Maximum Program Gain for All Sites
should be kept. For most sites, carefully examine the annotations, overhead assumptions, and related code.
Within the upper-right area of the
Suitability Report
window, if multiple parallel sites were detected during execution, select a different Site row to display its details.

Implementing Modeling Assumptions Later When Adding Parallel Code

In the lower-right part of the Suitability Report under
Runtime Modeling
, changing a checkmark does not resolve an issue - it configures the modeling of your proposed parallel program execution. To implement the modeled improvements indicated by the check boxes, use specific parallel framework constructs in the
Add Parallel Framework
step of the workflow and consider .