240
edits
Changes
→How do you actually use it?
6. Clean the Solution and Build the Project to generate the binary.
[[File:Gpu-project 1-6-l.png | left]][[File:Gpu-project 1-6-r.png | right]]
7. Launch Advisor through Windows Start > All Programs > Intel Parallel Studio XE 2017 > Analyzers > Advisor 2017
[[File:Gpu-project 1-12-t.png]]
[[File:Gpu-project 1-12-b.png]]
13. Check Collect information about FLOPS, L1 memory traffic, and AVX-512 mask usage for a complete Trip Count Analysis; this step is optional.
[[File:Gpu-project 1-13.png]]
14. Under the Binary/Symbol Search tab, add the visual studio project’s Release folder as a search directory. There will be warnings saying you are missing some symbols during the Survey Analysis, please ignore them.
[[File:Gpu-project 1-14.png]]
15. Under the Source Search tab, provide the location of the application’s source code.
[[File:Gpu-project 1-15.png]]
16. Select OK to complete the project creation process.
1. Allow Advisor to survey the application by clicking on the Collect button under the Threading Workflow tab (on the left panel).
[[File:Gpu-project 2-1.png]]
2. Continue profiling by running the Trip Counts and FLOPS analysis.
[[File:Gpu-project 2-2.png]]
''' Further Analysis: '''
2. To add annotations, include the <advisor-annotate.h> header file.
[[File:Gpu-project 3-2.png]]
3. Mark a possible parallel site and task with the following macros:
[[File:Gpu-project 3-3.png]]
Here are the syntax for the annotations:
ANNOTATE_SITE_BEGIN(Site 1);
5. Checking the checkboxes beside certain sites will mark them for deeper analyses.
[[File:Gpu-project 3-5.png]]
6. With one of the sites checked, run the Dependencies Analysis.
[[File:Gpu-project 3-6.png]]
7. For this example, there should be no dependencies. However, this is one warning: One task in parallel site; right click on the warning and select the What Should I Do Next? option.
NOTE: the What Should I Do Next? option is very useful for opening the documentations on the module you are pointing at.
[[File:Gpu-project 3-7.png]]
8. Go back to the Survey Report and uncheck the Deeper Analysis checkbox beside the target site.
9. Once you have annotated the sites and their tasks, run the Suitability Analysis.
[[File:Gpu-project 3-9.png]]
10. Since OpenMP is the focus of this workshop, change the Threading Model to OpenMP. Next, set the CPU Count to the amount of processors available on the machine.
[[File:Gpu-project 3-10.png]]
11. Load Imbalance and Runtime Overhead will change as you modify the Avg. Number of Iterations (Tasks) and the Avg. Iteration (Task) Duration sliders and click Apply.
[[File:Gpu-project 3-11.png]]
12. Estimated performance will also increase if you check the Runtime Modeling checkboxes that have benefits attached. The blue links will explain the means to enable the enhancements.
[[File:Gpu-project 3-12.png]]
=== Resources ===
