forum@abinit.org
Subject: The ABINIT Users Mailing List ( CLOSED )
List archive
Re: PARALLEL BUG Re: Re: [abinit-forum] error during nonlinear rf on the cluster
Chronological Thread
- From: "Anglade Pierre-Matthieu" <anglade@gmail.com>
- To: forum@abinit.org
- Subject: Re: PARALLEL BUG Re: Re: [abinit-forum] error during nonlinear rf on the cluster
- Date: Wed, 2 Aug 2006 11:59:36 +0200
- Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=PmsxZICEekoC/8C1dn/LnTn9HunhmMdXpKiTlX3gkJbSoR8VvQZaNC/yv8yQg00k5+uQeLXGpsOnGgbHqCyuqmYauQKSe/+A1ZrO0+vYbHg9N8Ox4bjEIgzP4jbA339/LqlF09bJ5nfQzomhR40sf1XhkwOFq1cdq9tQXkQkrCU=
Hi,
I think I won't be able to reproduce this bug. The version of abinit I
use, thought a standard 4.6.5, is not able to go as far as yours is
going. Sorry.
PMA
On 8/2/06, ilukacevic@fizika.unios.hr <ilukacevic@fizika.unios.hr> wrote:
In the first place I am apologizing for the too long post.
Regarding kptopt: I followed the Tutorial input file tnlo_2.in. Dataset1
should consider k points in the irr. BZ (kptopt=1), and Dataset2 in the full
BZ (kptopt=2; although, kptopt=3 considers full BZ, and kptopt=2 only half of
the BZ - this is not fully understandable to me).
Regarding parallel comp. flag: I am not sure of this, because of the previous
calculation I have made. I am sending a makefile_macros from the cluster as a
reference anyway.
Regarding original values: I wouldn't mind at all. First I used ngkpt=4 4 4
and ecut=50. After the message (which I posted previously) I had increased
ecut to 60. Again I got the same massage. Then I increased ngkpt to 8 8 8
(keeping ecut=60) and then the calculation stoped with an error.
As a reference I am also sending my anaddb output file for the case ngkpt=6 6
6 and ecut=60.
Thank You very much for troubling with this.
I. Lukacevic
makefile_macros file:
# For the sequential version
# Machine type
MACHINE=P6
# Fortran optimized compilation
FC=/opt/mpich/intel/bin/mpif90
FFLAGS = -O3 -w -tpp7 -axW -static-libcxa
FFLAGS_Src_5common = -O2 -mp -xW -tpp7 -static-libcxa
FFLAGS_Src_2psp = -O0 -xW -tpp7 -static-libcxa
FFLAGS_Src_2nonlocal = -O0 -xW -tpp7 -static-libcxa
#FFLAGS_Src_6response = -O0 -xW -tpp7 -static-libcxa # not effective for
Test_v4
FFLAGS_LIBS= -O3 -w -static-libcxa
LIBS=$(LAPACK) $(BLAS) -L/opt/intel/fc/9.0/lib -L/opt/mpich/intel/lib -lsvml
-lcxa -static-libcxa /opt/intel/cmkl/8.0/lib/32/libguide.a
/opt/intel/cmkl/8.0/lib/32/libmkl_def.so
/opt/intel/cmkl/8.0/lib/32/libmkl_p4.so
/opt/intel/cmkl/8.0/lib/32/libmkl_lapack.a -lunwind -lguide
# C preprocessor, used to preprocess the fortran source.
CPP=/lib/cpp
CPP_FLAGS=-P -traditional -D__IFC
# The cpp directive CHGSTDIO changes the standard I/O definition
# Uncomment the next line for this to happen.
#CPP_FLAGS=-P -DCHGSTDIO
# C optimized compilation.
CC=/opt/mpich/intel/bin/mpicc
CFLAGS=-O
# Location of perl. Used to generate the script fldiff, in ~ABINIT/Utilities.
PERL=/usr/bin/perl
# List of machine-dependent routines
MACHINE_DEP_C_SEQ_SUBS_LIST=etime.o
- PARALLEL BUG Re: Re: [abinit-forum] error during nonlinear rf on the cluster, Anglade Pierre-Matthieu, 08/01/2006
- <Possible follow-up(s)>
- Re: PARALLEL BUG Re: Re: [abinit-forum] error during nonlinear rf on the cluster, ilukacevic, 08/02/2006
- Re: PARALLEL BUG Re: Re: [abinit-forum] error during nonlinear rf on the cluster, Anglade Pierre-Matthieu, 08/02/2006
- Re: Re: PARALLEL BUG Re: Re: [abinit-forum] error during nonlinear rf on the cluster, ilukacevic, 08/02/2006
Archive powered by MHonArc 2.6.16.