[CPMD-list] SIGNAL: Floating point exception

[Juerg Hutter]

Hi

your problem is most likely related to the Nose thermostat
for the electrons. The integrators are not able to handle
your system and you get diverging values, depending on the
cutoff this happens sooner or later during the job.
The job stops in another part of the program because
only then the large numbers cause an exception.

This type of problem can usually be handled by adjusting
the degrees of freedom for the thermostat. 
You have to change the 4th parameter in the 
NOSE PARAMETER list and it is probably also a good idea to
crank up the integrators (parameter 5 and 6).
Have a look in the manual at the HINTS section.

BTW: NPT simulations for an isolated molecule, I assume
this is only a test.

Juerg


----------------------------------------------------------
Juerg Hutter                   Phone : ++41 1 635 4491
Physical Chemistry Institute   FAX   : ++41 1 635 6838
University of Zurich           E-mail: hutter at pci.unizh.ch
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland
----------------------------------------------------------


On Wed, 17 Apr 2002, Martijn Zwijnenburg wrote:

> Hi,
> 
> When I try to perform a MD run on a isolated molecule (29 atoms, 
> of which 10 H, NPT ensemble) on 8 processors of a Cray T-3E I 
> get the following error message:
> 
> SIGNAL: Floating point exception (floating point overflow)
> 
>  Beginning of Traceback (PE 0):
>   Interrupt at address 0x80053a3e8 in routine 'SGEMM@'.
>   Called from line 158 (address 0x80019d480) in routine 'RORTHO'.
>   Called from line 35 (address 0x80019c8fc) in routine 'RORTOG'.
>   Called from line 119 (address 0x80019aef4) in routine 'POSUPA'.
>   Called from line 211 (address 0x80017c03c) in routine 'MDMAIN'.
>   Called from line 123 (address 0x800179cf4) in routine 'MDPT'.
>   Called from line 114 (address 0x800001774) in routine 'CPMD'.
>   Called from line 3 (address 0x80000141c) in routine 
> 'CPMD_STUTTGART'.
>   Called from line 475 (address 0x800000c94) in routine '$START$'.
>  End of Traceback.
> Floating exception (core dumped)
> 
> I notticed that there is an influence of the cutoff. At a cutoff of 80 it 
> runs 10 steps before crashing, while at 40 it runs 327 steps before 
> crashing with the same error message. 
> What could be the reason for this floating point overflow? The 
> influence of the cutoff sugests something with lack of memory but I 
> can not combibe that with a floating poitn overflow. Anyway, all 
> sugestions are welcome.
> 
> Regards,
> 
> Martijn Zwijnenburg
> 
> -------------------------------------------------------------------------
> Martijn Zwijnenburg
> Lab. of Applied Organic Chemistry and Catalysis
> Delft University of Technology
> Julianalaan 136
> 2628 BL Delft
> The Netherlands
> Tel: 0031-(0)152782691
> Fax: 0031-(0)152784700
> e-mail: M.A.Zwijnenburg at tnw.tudelft.nl
> web page: http://come.to/tock
> _______________________________________________
> CPMD-list mailing list
> CPMD-list at cpmd.org
> http://www.cpmd.org/mailman/listinfo/cpmd-list
>