Performance considerations

While the calculations performed by FLOATSOFT itself are rather simple, interfacing with SOLIDWORKS API and the necessity to use it’s Virtual Bodies is the limiting factor of FLOATSOFT performance. Some SOLIDWORKS models will work faster than others; FLOATSOFT Settings play a major role as well. The learning algorithm attempts to minimize the necessity to interface with SOLIDWORKS API, and improve the performance. While the initial few calculations might seem sluggish, each subsequent calculation will become faster with that specific model. Here are some guidelines to reduce calculation times even more:

• Use a CPU with strong single-thread performance (you can check it here).
• Use a SOLIDWORKS model with simpler shapes and less SOLIDWORKS Features. If complexity cannot be avoided, export the hull as a Body to a separate .sldprt file to reduce the active SOLIDWORKS Feature count.
• Set the Error margin in the Settings window to a higher value. While this obviously reduces the accuracy of the calculations, you might find it preferable during the initial, ball-park design phase.
• When using Iterative modes, increase the Range Step size in the Iteration Range area.

Note: reducing or increasing the number of Decimal Points in the Settings window does not affect the performance.