[Pw_forum] Running parallel in some example of 2 cup is wrong?

[lucking-pine]

Hi,
Thank you for Duy Le and GAO Zhe's help.I will follow your suggestion to try
again.

在 2011年5月27日 上午12:15，GAO Zhe <flux_ray12 at 163.com>写道：

> In example 03, the first step is a calculation of CO at GAMMA point.
> Thus, you cannot use more than one pool to run it. Why not try *mpiexec -n
> 8 pw.x -npool 1* and see how it goes?
> --
> GAO Zhe
> CMC Lab, MSE, SNU, Seoul, S.Korea
>
> At 2011-05-26 22:35:08，lucking-pine <songsong19840614 at gmail.com> wrote:
>
> My computer is 2 cpu and 8 core of Dell. In some examples are crash. Such
> as:co.out of example 03 show that 'from d_&_i:error;some nodes have no k
> points'.
> And the environment_varibles of examples I set :
>
> PARA_PREFIX="mpiexec -n 8"
> PARA_POSTFIX="-npool 2"
>
> And how odd is when I set :
>
> PARA_PREFIX="mpiexec -n 4"
> PARA_POSTFIX="-npool 1"
> the example 03 can be work.
>
> The similar case occur also in some of other examples.
>
>  Please tell what problem is…………thanks.
> ps:my compiler is:icc11.1.075,ifort11.1.075,mpich2,pwscf 4.2.1
> system is red hat enterprise linux 5
>
> Best
> Pine.
>
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20110527/14bbb538/attachment.html>