[Pw_forum] (Slightly) different energies based on ordering of atomic positions

[Paolo Giannozzi]

Mike Mehl wrote:

> I suspect that the problem is that the first non-symmorphic symmetry 
> operation found is different in the two runs, so different space group 
> operations are thrown out of the two calculations.

no, I don't think this is the problem. In both cases the code just
discards fractional translations. It should find the same symmetry
operations (this can be verified using "verbosity='high'"). Note
that the first scf iterations give exactly the same results, then
a small difference starts to build up. Why? hard to say.

> I find it a bit disconcerting to get two different answers
> from what should be identical input.

me too. However, once you have spent several days chasing a bug
that turns out to be due to different processors producing
different results from the same operations on the same data
(it happened to me some time ago on a BlueGene), you abandon
any hope to get reproducible numbers with more than a few
significant digits.

Paolo
-- 
Paolo Giannozzi, Democritos and University of Udine, Italy