Intel® Advisor User Guide

User Guide

Contents

Run
Offload Modeling
 Perspective

Select the
Offload Modeling
 perspective to identify:
  • If offloading your code to a target device provide benefits and what is the potential speed-up
  • What can potentially limit performance of your offloaded application
  • What code regions are recommended for offloading
  • What code regions are not recommended for offloading and why
  • How effectively data is transferred between an original and the target devices
To run the
Offload Modeling
 perspective:
  1. Choose collection accuracy level to select perspective steps and set analysis properties, depending on the desired results:
    • Low
      : Model your application performance for a target device and get the basic information about potential speed-up and performance.
    • Medium
      : Model your application performance and data transfers between host and target devices.
    • High
      : Model your application performance and data transfers and analyze dependencies to improve offload modeling accuracy.
    • Custom
      : Customize the perspective flow and properties.
    The higher accuracy value you choose, the higher runtime overhead is added to your application. The
    Overhead
     indicator shows the overhead for the selected configuration. If you select
    Custom
     accuracy, the overhead is calculated for selected analyses and properties.
    By default, accuracy is set to Low. For more info, see Offload Modeling Accuracy Presets.
  2. If you want to check the offload profitability for specific loops/functions instead of analyzing the whole application:
    1. Go to
      Project Properties
      Performance Modeling
      .
    2. Enter the
      --select
      =
      [(r|recursive):]<id>|<file>:<line>|<criteria>[,<id>|<file>:<line>|<criteria>,..]
       in the
      Other parameters
       field.
  3. Select a target platform from the drop-down.
  4. Run the perspective: click button.
    While the perspective is running, you can do the following in the
    Analysis Workflow
     tab:
    • Control the perspective execution:
      • Stop data collection and see the already collected data: Click the button.
      • Pause data collection: Click the button.
      • Cancel data collection and discard the collected data: Click the button.
    • Expand an analysis with to control the analysis execution:
      • Pause analysis and see the already collected data: Click the button.
      • Stop analysis and start the next analysis selected: Click the button.
      • Interrupt execution of all selected analyses and see the already collected data: Click the button.
    You can generate command lines for selected perspective configuration by clicking the
    Command Line
     button. For CLI wokflow example, see Command Line Use Cases.
When the data is collected,
Intel® Advisor
 generates the
Offload Modeling
 report. Depending on the selected accuracy level and perspective properties, continue to investigate the results:
After you run the
Offload Modeling
 perspective, the collected Survey data becomes available for all other perspectives. If you switch to another perspective, you can skip the Survey step and run only perspective-specific analyses.

Product and Performance Information

1

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.