User Guide


Enforce Offloading for Specific Loops

Model performance on a target device for specific loops only even if they are not profitable.
If you want to check offload profitability only for specific loops or if your loop of interest is reported as not recommended for offloading to an accelerator, you can model performance only for these loops.
: In the commands below, make sure to replace the
with your application executable path and name
executing a command. If your application requires additional command line options, add them
the executable name.
To do this:
  1. Collect performance data using the
    script or the
    command line interface. For example:
    advisor --collect=survey --static-instruction-mix --project-dir=./advi_results -- ./myApplication
    advisor --collect=tripcounts --flop --enable-cache-simulation --target-device=gen11_icl --project-dir=./advi_results -- ./myApplication
  2. Rerun performance modeling using the
    advisor-python $APM/ ./advi_results --select-loops=foo.cpp:34,bar.cpp:192 --enforce-offload
    Make sure to use the following options:
    • --select-loops=
      to specify loop(s) of interest in a comma-separated list
    • --enforce-offloads
      to make sure all of them are offloaded
    • Optional
      to ignore loop-carried dependencies when estimating offload profitability for specific loops
      Specify the same loops/functions or a subset of loops/functions specified with the
Open the
Offload Modeling
results in the
Intel® Advisor
GUI or view an HTML report. The results will show performance modeled for the selected loops only.

Product and Performance Information


Performance varies by use, configuration and other factors. Learn more at