Aim of the Performance Analyzer

There are many possible reasons why the simulation of a model can be slowed down, and it is not always straightforward to understand how to improve the performance of the solver. This difficulty often arises in situations such as:

• Parameters which are not set properly.
• Very large models for which it is not easy to understand causes and effects.
• Models which suddenly slow down when they were running fine an instant ago.
• A lack of modeling experience.

The total time taken by the solver to simulate the model depends on factors such as the complexity of the model and the solver characteristics, including:

• Number and type of equations being solved.
• Number of discontinuities.
• Speed (or frequency) at which variables are varying in the model.
• Non-linearity of the equations.
• Solver settings, integration step size and accuracy.
• User submodels improperly coded.
• Hardware and network infrastructure, including computer speed.
• Software configuration, especially the compiler used to generate the model executable.

Solver and equation terminology is explained throughout the manual, and the Glossary at the end of the manual provides more detailed definitions.

The aim of the Performance Analyzer is to gather information about the model and about the solving process in order to assess how the solving performance can be improved. The provided information covers most of the points listed above; interpreting this information correctly can lead to significant performance improvements.