From: Cade Markegard (cmarkega_at_uci.edu)
Date: Fri Jul 06 2012 - 13:56:42 CDT
Robert,
I should have posted this up, but you need to download the source code and
compile NAMD on the cluster you are going to use it on. Each cluster has
their own MPI libraries, so a general build of NAMD posted won't work for
this case. I just followed the instructions in the notes.txt file of the
source code's directory for building with MPI. Post up here if you still
have any questions, I'll be happy to help you out with it.
Best of Luck,
Cade Markegard
UC Irvine Chemical Engineering Graduate Student
http://compnano.eng.uci.edu
On Fri, Jul 6, 2012 at 11:40 AM, Robert Elder <rmelder_at_gmail.com> wrote:
> Hi all,
>
> I am having the same issue with replica exchange simulations as the user
> below, where using the +replicas flag throws the error:
>
> FATAL ERROR: Unknown command-line option +replicas
>
> I'm using the latest NAMD2.9 binary and simply attempting to run on my
> local workstation, not a cluster as with the user below. However, I would
> like to run REMD simulations on a Cray XT5 (Kraken), so I would prefer to
> use the MPI version rather than the TCL socket version of REMD.
>
> I can provide any details that seem relevant. Your help is appreciated!
>
> Thanks,
> Robert Elder
> CU Boulder Chemical Engineering Graduate Student
>
> On Mon, Jun 4, 2012 at 3:57 PM, Cade Markegard <cmarkega_at_uci.edu> wrote:
>
>> Hi All,
>>
>> I am currently trying to run the NAMD 2.9 replica exchange example. I am
>> on a Rocks Cluster (v. 5.4.3) with all the rolls installed, and the NAMD
>> binary I am using is from NAMD_2.9_Linux-x86_64. When I run the example
>> command located in fold_alanin.conf but without the hostfile option (since
>> I am trying to run this on a single node):
>>
>> mpirun -np 8 $bindir/namd2 +replicas 8 job0.conf +stdout
>> output/%d/job0.%d.log
>>
>> where $bindir is my NAMD 2.9 path, I am currently receiving the error:
>>
>> *FATAL ERROR: Unknown command-line option +replicas*
>>
>> Is this the wrong pre-compiled NAMD binary or is there something else
>> that I am missing here? Could someone please point me to the correct
>> binary? I rather not have to compile NAMD from the source... Also I really
>> like the idea of the new MPI based method implemented in 2.9 rather than
>> the TCL method in 2.8, so when it comes time it will be easier to run a
>> simulation across multiple nodes.
>>
>> Any help regarding this subject would be greatly appreciated!
>>
>> Many thanks,
>> Cade Markegard
>> UC Irvine Chemical Engineering Graduate Student
>> http://compnano.eng.uci.edu
>>
>
>
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:45 CST