# Precision Limits

When selecting a system of units, be careful to avoid calculations that approach the precision limits
of the computer hardware. For most modern computers, the range is approximately 10^{-300} to 10^{300} in double-precision.
If numbers exceed these limits, the program may produce artifacts in the model that may be difficult to identify or detect.

In addition, all calculations are performed using double-precision floating-point values. This provides approximately 14 digits of precision. If the coordinates of your model are offset by a large value, this can remove effective digits of precision from your calculation. For this reason, we recommend that the model coordinate system be offset so that the origin is near or inside the model extent.

Was this helpful? ... | FLAC3D © 2019, Itasca | Updated: Feb 23, 2021 |