Re: Can any one help me with this parallel issue

From: Sterling Paramore (paramore_at_hec.utah.edu)
Date: Mon Jan 30 2006 - 10:02:17 CST

The first problem I see is that you're using the AMD64 compiling options
when the Xeon is an intel processor. I'm not sure if that will make
much of a difference, but you should use the i686 options instead. If
you have the intel compiler, you should definitely use that. I believe
you can download an academic copy for free. Here's what I did on the
Xeon EM64T I have access to (since you're not using myrinet, you don't
need the gm library):

CHARMM:

./build charm++ mpi-linux icc gm2 --no-shared -DCMK_OPTIMIZE=1
-I/opt/mpi/x86_64-intel-8.1/mpich-gm-1.2.6..14a/include
-L/opt/mpi/x86_64-intel-8.1/mpich-gm-1.2.6..14a/lib -lgm -lmpich
-lpmpich -lpthread

FFTW:

/configure --enable-float --enable-type-prefix
--prefix=/home/airforce/paramore/NAMD_2.6b1_Source/fftw/linux-arl

TCL:

./configure
--prefix=/home/airforce/paramore/NAMD_2.6b1_Source/tcl/linux-arl
--disable-shared

NAMD:

edited arch/Linux-i686-icc.arch and arch/Linux-i686-icc-GM.arch to point
to right charmm compilation

./config tcl fftw plugins Linux-i686-GM-icc

-Sterling

Morad Alawneh wrote:

> *Dear everyone,
>
> I have recently compiled namd2.6b1 to have it work in parallel. I have
> a system of about 92,000 atoms, and when I submit my job using more
> than 2 processors I get bad performance, which is lower than using 2
> processors.
> Would anyone help me to figure out what is the problem.
>
> Thanks
>
> Cluster system:
> It is a Dell 1855 Linux cluster consisting of 630 nodes. Each node is
> equipped with two Intel Xeon EM64T processors (3.6GHz) and 4 GB of
> memory. Each node runs its own Linux kernel (Red Hat Enterprise Linux
> 3). All nodes are connected with Gigabit Ethernet
>
> Installation instructions:
> _charm++ installation:_
>
> Edit namd/Make.charm
> (set CHARMBASE to the full path to charm-5.9)
>
> Edit the file charm/src/arch/mpi-linux-amd64/conv-mach.sh to have the
> right
> path to mpicc and mpiCC (replace mpiCC by mpicxx), use mpich/gnu.
>
> ./build charm++ mpi-linux-amd64 gcc --libdir=/opt/mpich/gnu/lib
> --incdir=/opt/mpich/gnu/include --no-build-shared -O -verbose
> -DCMK_OPTIMIZE=1
>
>
> _namd Installation:_
>
> Edit various configuration files:
> namd/arch/Linux-amd64-MPI.arch (fix CHARMARCH to be mpi-linux-amd64-gcc)
> namd/arch/Linux-amd64.fftw (fix path to files, and delete all
> -I$(HOME)/fftw/include -L$(HOME)/fftw/lib)
> namd/arch/Linux-amd64.tcl (fix path to files, and delete all
> -I$(HOME)/tcl/include -L$(HOME)/tcl/lib)
>
> Set up build directory and compile:
> ./config tcl fftw Linux-amd64-MPI
> cd Linux-amd64-MPI
> make
>
>
> Job submission script:
> *
>
>*#!/bin/bash
>
>#PBS -l nodes=1:ppn=2,walltime=336:00:00
>#PBS -N Huge_gA
>#PBS -m abe
>#PBS -M alawneh_at_chem.byu.edu
>
># The maximum memory allocation
>let Memory=256*1024*1024
>export P4_GLOBMEMSIZE=$Memory
>
>#cd into the directory where I typed qsub
>cd $PBS_O_WORKDIR
>
>TMPDIR=/ibrix/scr/$PBS_JOBID
>PROG=/ibrix/apps/biophysics/namd/Linux-amd64-MPI/namd2
>ARGS="+strategy USE_GRID"
>IFILE="equil1_sys.namd"
>OFILE="equil1_sys.log"
>
># NP should always be: nodes*ppn from #PBS -l above
>let NP=1*2
>
># Full path to mpirun
>MPIRUN=/opt/mpich/gnu/bin/mpirun
>
># Execute the job using mpirun
>$MPIRUN -machinefile $PBS_NODEFILE -np $NP $PROG $ARGS $IFILE > $OFILE
>
>exit 0*
>
> --
>
>
>
> /*Morad Alawneh*/
>
> *Department of Chemistry and Biochemistry*
>
> *C100 BNSN, BYU*
>
> *Provo, UT 84602*
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:43:16 CST