|
|
 |
|
|
Download:
batch_run_simple.sh
The script will run your serial (non-parallel) code from the current working directory.
Easy to modify it for your case since there is no need to keep track with
the input and output files copying to and from the sctarch run space as in
the next script, batch_run.sh. However, it may give you a very poor performance, especially
in case of intensieve I/O, as it utilizes a network file system (NFS), which is
much slower than a local disk.
Download:
batch_run.sh
The script will run your serial (non-parallel) code from a local /tmp
scratch directory on a compute node. You need to specify the executables
as well as the input and output files it has to copy to and from the scratch space, respectively.
Download:
mpi_run.sh
The scripts runs a parallel MPI job from your current working directory.
Download:
mpich2_gnu_script.sh
The script runs MPICH 2 parallel job from your current working directory. It implies MPICH 2 used with GNU (GCC and GFORTRAN) compilers.
Download:
mpich2_pgi_script.sh
The script runs MPICH 2 parallel job from your current working directory. It imp
lies MPICH 2 used with PGI compilers.
Download:
sander_mpi.sh
The script runs MPI-parallel sander (AMBER) from your current working directory.
Download:
namd_mpi.sh
The script runs MPI-parallel NAMD from your current working directory.
Download:
gromacs_run_gcc.sh
The script runs MPI-parallel GROMACS from your current working directory.
Download:
batch_matlab.sh
The script runs Matlab in batch mode on the cluster.
Download:
batch_fluent.sh
The script runs Fluent in batch mode on the cluster. Note, it is
necessary to request the license complex in the script, options
-hard and -l
To the top
|
© 2003 All Rights Reserved.
Rutgers, The State University of New Jersey.
|
|
