[molpro-user] numerical result out of range

[Glen Jenness]

Dear Molpro Users,
Recently I tried performing a large calculation (75 atoms with ~1750
contracted basis functions) using the LBE0 functional (basically PBE0 with
the HF replaced with LHF), and encountered the following error:

Last System Error Message from Task 0:: Numerical result out of range

The calculation crashes right around the first iteration, and any
information that is printed looks perfectly normal and in line with what I
would expect.  Furthermore, I've ran this calculation in both serial and
parallel, and it only crashes in parallel (I've ran it on both 4 and 2
processors on two different computers with the same results).  However
when I run the job in serial, it works just fine (and the first iteration
results match up), so this seems to be a problem with parallel execution.

I've looked through the archives and saw a similar issue cropped up a
couple years back, but no solution appears to have been posted.  Any
suggestions would be welcome as to the source of this problem.

Thanks,
Glen Jenness

Jordan Group
Department of Chemistry
University of Pittsburgh