[david_1995]

Does anyone know what happens to the accuracy of the field functions when a simulation that has been run in a double precision version of the code is later opened (for viewing the solution only) in the mixed precision version? I am running simulations in a cluster which has the double precision version, but I view the solution on my workstation which has the mixed precision version.

[ping]

if you can see the difference between 7 digits of precision and 15 digits in your typical field function displays and plots then that would be quite amazing. dp is only important during solution for convergence reasons when using more advanced physics