[Openmd-users] Update

Dan Gezelter gezelter at nd.edu
Fri Sep 3 13:47:17 EDT 2010


Brett,

  I'm having trouble reproducing this here with the files you sent.  Any chance you can give me details on OS version, compiler version, MPI version, etc.?  I can try on a couple of more architectures, but I want to get as close as possible to what you are running on.

 --Dan

On Sep 1, 2010, at 11:14 AM, Brett Donovan wrote:

> Hi Dan
> 
> These are what I am using. Let me know if you want more details on the
> actual machines/compilers etc etc. As far as I know these are the gcc
> compiled compiled libraries. not intel. Thanks.
> 
> Best wishes
> 
> Brett
> 
> On Wed, Sep 1, 2010 at 1:33 PM, Dan Gezelter <gezelter at nd.edu> wrote:
>> Brett,
>> 
>>  Can you send me the md / frc files you used for this?  It sounds like a bug in the computation of the conserved quantity, although it should be using the MPI-safe routines from SimInfo and Thermo for the necessary quantities.   Why it would pop up in NPTxyz and not the other integrators is a bit of a mystery.  Do you also get this problem if you use NPTi?
>> 
>>  --Dan
>> 
>> 
>> On Aug 31, 2010, at 6:44 AM, Brett Donovan wrote:
>> 
>>> No problem, they start pretty early over there, many have another
>>> month here or so before they come back.
>>> 
>>> Just noticed a little feature here is that when running on our server
>>> [this time with the new update], I get the following difference
>>> between running in a single processor and on the MPI version.
>>> 
>>> The single processor [non MPI] runs whilst the MPI version gives the following:
>>> 
>>> StatWriter detected a numerical error writing: Conserved Quantity
>>> 
>>> Any ideas what this might be? Its on the NPTxyz ensemble that this is
>>> occuring - where I have changed over from NVT to NPTxyz
>>> 
>>> Thanks
>>> 
>>> Brett
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> On Fri, Aug 27, 2010 at 7:09 PM, Dan Gezelter <gezelter at nd.edu> wrote:
>>>> Brett,
>>>> 
>>>>  Sorry I've been out of touch lately.  The freshman chemistry students have descended en masse, and I'm still a little swamped.
>>>> 
>>>>  --JDG
>>>> 
>>>> On Aug 27, 2010, at 10:36 AM, Brett Donovan wrote:
>>>> 
>>>>> Hi Dan
>>>>> 
>>>>> I think we might have sourced the problems with the different versions
>>>>> of OMD running on different machines. I think the issues of the
>>>>> energies blowing up was possibly down to the code not updating on the
>>>>> super computing cluster and us running an old copy. I am told by the
>>>>> guys in charge of the cluster they were using the tarballs [not the
>>>>> SVN] and that they might not have been the latest version to update
>>>>> from. They have now updated and built from the SVN copy and that seems
>>>>> all good.
>>>>> 
>>>>> I cannot seem to recreate any issues with the sin term in the cosine
>>>>> bend, and my calculation from my development version does not differ
>>>>> so hopefully...all is well and the original implementation should hold
>>>>> well.
>>>>> 
>>>>> Best wishes
>>>>> 
>>>>> 
>>>>> Brett
>>>>> 
>>>>> Have a great weekend!
>>>> 
>>>> ***********************************************
>>>>  J. Daniel Gezelter
>>>>  Associate Professor of Chemistry
>>>>  Director of Graduate Admissions
>>>>  Department of Chemistry and Biochemistry
>>>>  251 Nieuwland Science Hall
>>>>  University of Notre Dame
>>>>  Notre Dame, IN 46556-5670
>>>> 
>>>>  phone:  +1 (574) 631-7595
>>>>  fax:    +1 (574) 631-6652
>>>>  e-mail: gezelter at nd.edu
>>>>  web:    http://www.nd.edu/~gezelter
>>>> ************************************************
>>>> 
>>>> 
>>>> 
>> 
>> ***********************************************
>>  J. Daniel Gezelter
>>  Associate Professor of Chemistry
>>  Director of Graduate Admissions
>>  Department of Chemistry and Biochemistry
>>  251 Nieuwland Science Hall
>>  University of Notre Dame
>>  Notre Dame, IN 46556-5670
>> 
>>  phone:  +1 (574) 631-7595
>>  fax:    +1 (574) 631-6652
>>  e-mail: gezelter at nd.edu
>>  web:    http://www.nd.edu/~gezelter
>> ************************************************
>> 
>> 
>> 
> <Membrane_Therm.md><FF_Dipole.frc>

***********************************************
  J. Daniel Gezelter
  Associate Professor of Chemistry
  Director of Graduate Admissions
  Department of Chemistry and Biochemistry
  251 Nieuwland Science Hall
  University of Notre Dame
  Notre Dame, IN 46556-5670

  phone:  +1 (574) 631-7595
  fax:    +1 (574) 631-6652
  e-mail: gezelter at nd.edu
  web:    http://www.nd.edu/~gezelter
************************************************





More information about the Openmd-users mailing list