Re: [AMBER-Developers] incompatible processor requirements in parallel tests

From: Jason Swails <jason.swails.gmail.com>
Date: Thu, 18 Feb 2010 15:17:18 -0500

On Thu, Feb 18, 2010 at 3:07 PM, Volodymyr Babin <vbabin.ncsu.edu> wrote:
>> I checked in the new rules to Makefile.  People can feel free to
>> remove it if they think it's pointless.  But I agree with Lachele that
>> there should probably be some documentation about it (at least some
>> blurb in the README).
>
> The rules are very convenient indeed. Another option (less flexible,
> but potentially more automatic) would be to set -np inside the Makefile
>
> sh$ export DO_PARALLEL='command -np AMBER_NTHREADS'

The problem with this is that not all MPIs use the same flag to
represent the number of threads. For instance, mpiexec for many MPIs
uses -n instead. -c is another common option. I don't think there's
any way of generalizing this to any MPI a user might have installed.
We could, of course, always make it only compliant with the included
MPIs (openMPI and lam) that can be built with amber. What trade-offs
for automation/flexibility should we make? I don't think there's any
'right' answer.

Another thing about the rules, is DO_PARALLEL still has to be set to
use the right number of processors, or every single test will give the
message 'need 4 threads', etc. I also may not have caught all of
them, since I only added the ones that printed messages to stdout
during the course of the test.

Thanks!
Jason

>
> in the Makefile:
>
> user_DO_PARALLEL="$DO_PARALLEL"
>
> for n in 2 4 8 16;
> do
>  export DO_PARALLEL=`echo $user_DO_PARALLEL | sed -e "s/AMBER_NTHREADS/$n/"`
>  echo "Using DO_PARALLEL=$DO_PARALLEL"
>  <run tests>
> done
>
> Best,
>
> Volodymyr
>
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>



-- 
---------------------------------------
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Thu Feb 18 2010 - 12:30:04 PST
Custom Search