Compare CPU Roofline Results
Use the
Roofline Compare
functionality to display Roofline chart data from other
Intel® Advisor
results or non-archived snapshots for comparison purposes to track optimization progress.
Prerequisites
To compare the GPU Roofline results, make sure to get the following:
- A baseline GPU Rooflineresultorsnapshot
- One or more GPU Roofline results or snapshots of thesameapplication with an optimization applied
To compare the results:
- Open a baseline GPU Roofline result/snapshot.
- From theComparedrop-down toolbar, click+to load a comparison result/snapshot. You can load multiple results/snapshots for comparison one by one.
When the comparison is uploaded:
- The filenames for uploaded results/snapshots are displayed in theCompared Resultsregion.
- Similar loops/functions from all compared results are recognized automatically. They are connected with a dashed arrow line. The performance improvement between the loops/functions is shown above the line, in per cent. The improvement is calculated as thedifferencein FLOPS, INTOPS, or OPS and Total Time.The arrows showing the relationship among loops/functions do not reappear if you upload a new comparison file.
- Loops from different snapshots are shown as different icons on the chart. For example, on the picture below, the baseline loops are shown as circles and comparison loops are triangles and diamonds.
- To highlight all dots from a specific compared result, open theComparedrop-down and hover over the result name.
- Each time you change the Roofline configuration or filter the dots on the chart, the comparison is updated automatically.
- You can remove a selected result fromCompared Resultsby hovering over it and clicking theXicon. The result is removed from the chart and appears in theReady for comparisonregion. Click a name in theReady for comparisonregion to reload the result back to the chart.
- You can save the comparison itself to a file using the export feature.
To find the same loops/functions among the results,
Intel Advisor
compares several loop/function features, such as their type, nesting level, source code file name and line, and function name. When a certain threshold of similar or equal features is reached, the two loops/functions are considered a match and connected with a dashed line.
However, this method still has few limitations. Sometimes, there can be no match for the same loop/function if one is optimized, parallelized, or moved in the source code to four or more lines from the original place.
Intel Advisor
tries to ensure some balance between matching source code changes and false positives.