question

No. The 18.2 FlexScript Code Profiler’s units were not accurate. They’re only useful for comparing calls within a given profiling session to each other. Exporting it doesn't make sense because comparing one run of it to another isn't accurate.

The timing shown isn’t an actual measuring of the execution time. It is an estimate based on sampling of what is getting called and how long it might have been in that function relative to another function. You shouldn’t use those numbers as a measurement of how long it was in that function. You should only use it as a guideline for how long functions ran relative to other functions to help direct you to what functions might be running slowly in your model. The 18.2 FlexScript Code Profiler was also very limited in its scope; it records only FlexScript code execution; other processing is not taken into account. For a given model, unless you have some really critical, poorly optimized FlexScript functions, the likelihood is that your model processing speed is more dependent on other processing than on FlexScript code execution.

We replaced this profiler entirely in FlexSim 20.1.0 with a different tool that more accurately records how much time the CPU spends in profiled functions while the model runs. See https://docs.flexsim.com/en/20.1/Reference/Tools/PerformanceProfiler/PerformanceProfiler.html