Folks,
I am running a geometry optimisation (CG) of a molecule on a surface that takes
quite some time, using SIESTA 3.1 in parallel. It finally stops with an error:
New_DM. Step: 499
Re-using DM from previous geometry...
Re-using DM without extrapolation (not allowed)
Density Matrix sparsity pattern changed.
rank 0 in job 1 wardlaw122_52303 caused collective abort of all ranks
Since it's always at step 499, I assume that the program tries (and fails) to
reset or restart something in the DM code at optimisation step 500. Is there a
way to get around this?
I remember doing optimisations of 2000 steps with 2.0.2, so it must be a
relatively new feature that causes this problem.
Thanks in advance,
Herbert
--
Herbert Fruchtl
Senior Scientific Computing Officer
School of Chemistry, School of Mathematics and Statistics
University of St Andrews
--
The University of St Andrews is a charity registered in Scotland:
No SC013532
