[gmx-developers] Different results with Verlet scheme due to PME dipole correction

[David van der Spoel]

We've found some serious although uncommon errors in forces and energies 
when using the Verlet scheme:

- when using ewald_geometry = 3dc without vacuum layer in the box
- when using ewald_geometry = 3d (default) and epsilon_surface != 0

For more info see
http://redmine.gromacs.org/issues/1645

Please note that both these conditions are very unlikely to happen.

-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205.
spoel at xray.bmc.uu.se    http://folding.bmc.uu.se