Difference between revisions of "HowTo:nwchem"

From CAC Wiki
Jump to: navigation, search
(Gaussian)
(NWChem)
Line 1: Line 1:
 
= NWChem =
 
= NWChem =
  
This is a short "Frequently Asked Questions" file on using the parallel electronic-structure code "NWChem" on systems at the Centre for Advanced Computing. This software uses MPI as a message-passing system and is (in principle) able to run on an arbitrary number of processors. Its ability to perform a very broad spectrum of molecular-structure calculations, ranging from CI to ab-initio molecular dynamics, makes it an interesting alternative to the standard electronic structure code Gaussian.  
+
This is a short help file on using the parallel electronic-structure code "NWChem" on systems at the Centre for Advanced Computing. This software uses MPI as a message-passing system and is (in principle) able to run on an arbitrary number of processors. Its ability to perform a very broad spectrum of molecular-structure calculations, ranging from CI to ab-initio molecular dynamics, makes it an interesting alternative to the standard electronic structure code Gaussian.  
 
{|  style="border-spacing: 8px;"
 
{|  style="border-spacing: 8px;"
 
| valign="top" width="50%" style="padding:1em; border:1px solid #aaaaaa; background-color:#e1eaf1; border-radius:7px" |  
 
| valign="top" width="50%" style="padding:1em; border:1px solid #aaaaaa; background-color:#e1eaf1; border-radius:7px" |  
Line 29: Line 29:
 
<pre>use nwchem</pre>
 
<pre>use nwchem</pre>
 
which will set you up automatically.
 
which will set you up automatically.
 +
|}
 +
{|  style="border-spacing: 8px;"
 +
| valign="top" width="50%" style="padding:1em; border:1px solid #aaaaaa; background-color:#f7f7f7; border-radius:7px" |
  
== Scratch files ==
+
== Running NWChem from a command line==
  
One of the settings is the environment variable '''GAUSS_SCRDIR''' which is required to redirect the temporary files that Gaussian uses to the proper scratch space, presently
+
Like other electronic-structure programs, NWChem is run by supplying an input file that defines the system on which to perform a calculation (usually a molecule, or a group of molecules), and the method to use (i.e., the level of calculation, such as "Hartree-Fock", the basis set, and other details of the computation).
  
<pre>export GAUSS_SCRDIR=/scratch/hpcXXXX</pre>
+
The variety of possible calculations is great, and so is the complexity of systems. It is impossible for us here to explain the format that a NWChem input file needs to have. This is explained in the [http://www.nwchem-sw.org/index.php/Release62:NWChem_Documentation User's Manual] which is available online.
  
where hpcXXXX stands for your username. If for some reason Gaussian does not terminate normally (e.g. a job gets cancelled), it leaves behind large '''scratch files''' which you may have to delete manually. To check if such files exist, type
+
Here, we provide a simple sample input file which should be given the file extension .nw.
  
<pre>ls -lt /scratch/hpcXXXX</pre>
+
<pre>
 +
start h2o
  
Once you have determined that the scratch files are no longer needed (because the program that used them is not running any more), you can delete them by typing
+
title "H2O, cc-pVDZ basis, SCF optimized geometry"
  
<pre>rm /scratch/hpcxxxx/Gau-*</pre>
+
geometry units au
 +
H      0.0000000000  1.4140780900  -1.1031626600
 +
H      0.0000000000  -1.4140780900  -1.1031626600
 +
O      0.0000000000  0.0000000000  -0.0080100000
 +
end
  
Cleaning up the scratch space is the user's responsibility. If it is not done regularly, it can cause jobs to terminate, and much work to be lost.
+
basis
|}
+
H library cc-pVDZ
{|  style="border-spacing: 8px;"
+
O library cc-pVDZ
| valign="top" width="50%" style="padding:1em; border:1px solid #aaaaaa; background-color:#f7f7f7; border-radius:7px" |
+
end
== Running Gaussian from a command line==
+
  
To run Gaussian on our systems, you have to belong to a '''user group g98''' (it's called that for historical reason, but it applies to all versions of Gaussian). You need to read our license agreement and [http://www.hpcvl.org/sites/default/files/gaussian-statement.pdf signed a statement] to be included in this user group. Once you are, you can access the executables.
+
scf
 +
  thresh 1.0e-8
 +
end
  
A computation is performed by preparing an input file and pipe it to standard input of the program '''g09'''. Standard output should be caught in a log-file. We suggest you use the '''extensions''' ''.g09 for input'' files and ''.log for results''.
+
task scf
 +
</pre>
  
Interactively, the command line to invoke Gaussian is thus:
+
This extension may be omitted when calling the program. NWChem creates typically a whole array of output files that are documented in the User's Manual. A general log is displayed on the console, and may be saved in a file by redirecting the standard output:
  
<pre> g09 < test.g09 >test.gout </pre>
+
<pre>nwchem sample > sample.log</pre>
  
This will only work if you are a member of the g98 group and have set the environment correctly. Note that the '''interactive execution of Gaussian is only meant for test runs'''.
+
where we assume that your input file is called sample.nw and you want to save the log to a file sample.log.
  
Gaussian input files are explained in the "User's Reference". It is impossible to give an outline here. Sample files can be found in  
+
Note that this is just the basic syntax of the program call. In practise you will use a parallel environment to execute the program (see next section). In fact, executing NWChem by just typing the above line will run it in serial mode.
  
<pre>/opt/gaussian/g09/bsd/logs</pre>
+
== Parallel Runs ==
  
'''Note:''' It is absolutely essential to have a good idea about the size and complexity of your calculations before you start a Gaussian job. Many of the methods mentioned above have terrible scaling properties, i.e. the computational cost grows very quickly with the number of electrons, degrees of freedom, or number of basis functions used. We suggest you start with a small basis set and a cheap method, and then slowly increase those parameters.
+
NWChem is inherently parallelized and designed to scale well on a multi-processor machine or a cluster. The underlying parallel system is MPI (Message Passing Interface) which is a commonly available standard that runs on many platforms. Consult our [[HowTo:mpi MPI help file]] and follow some of the links in there if you want to have more information about MPI. Even if you want to use only one processor for your NWChem run (which sometimes is the best solution, particularly for smaller computations), you have to submit the program to a parallel environment. On our clusters, the relevant command is '''mpirun''':
 +
<pre>mpirun -np 8 nwchem sample > sample.log</pre>
 +
This will run your sample.nw input file on eight processors. Note that you are only allowed to run NWChem this way for small test systems! For any production jobs, you have to submit the task to the scheduler (see next section.).
  
 
== Submitting (parallel) Gaussian jobs ==
 
== Submitting (parallel) Gaussian jobs ==

Revision as of 17:39, 27 May 2016

NWChem

This is a short help file on using the parallel electronic-structure code "NWChem" on systems at the Centre for Advanced Computing. This software uses MPI as a message-passing system and is (in principle) able to run on an arbitrary number of processors. Its ability to perform a very broad spectrum of molecular-structure calculations, ranging from CI to ab-initio molecular dynamics, makes it an interesting alternative to the standard electronic structure code Gaussian.

Features

NWChem is an electronic-structure code that is suitable to perform complex calculations on molecular structure. It was specifically designed to perform well on high-performance parallel computers. The installation on the SunFire cluster of HPCVL employs the MPI message passing package for parallel execution.

NWChem allows, among others, the following calculations to be performed:

  • Hartree-Fock (e.g. RHF, UHF, ROHF etc.)
  • DFT including spin-orbit DFT, with many exchange and correlation functionals.
  • Complete Active Space SCF (CAS-SCF)
  • Coupled-Cluster (CCSD, CCSD+T, etc.)
  • limited CI (eg, CISD) with perturbative corrections
  • MP2 (2nd-order Mollar-Plesset Perturbation Theory)
  • In general: single-point calculations, geometry optimizations, vibrational analysis.
  • Static one-electron properties, densities, electrostatic potentials.
  • ONIOM model for multi-level calculations on larger systems
  • Relativistic corrections (Douglas-Kroll, Dyall-Dirac, spin-orbit)
  • Ab-initio molecular dynamics (Carr-Parinello)
  • Extended (solid-state) systems DFT
  • Classical force-fields (Molecular Mechanics: AMBER, CHARMM, etc.)

For a more complete list, see the official NWChem homepage and click on "capabilities".

Location of the program and setup

The NWChem program is located in the directory /opt/nwchem/bin. To access it, you have to use the usepackage command

use nwchem

which will set you up automatically.

Running NWChem from a command line

Like other electronic-structure programs, NWChem is run by supplying an input file that defines the system on which to perform a calculation (usually a molecule, or a group of molecules), and the method to use (i.e., the level of calculation, such as "Hartree-Fock", the basis set, and other details of the computation).

The variety of possible calculations is great, and so is the complexity of systems. It is impossible for us here to explain the format that a NWChem input file needs to have. This is explained in the User's Manual which is available online.

Here, we provide a simple sample input file which should be given the file extension .nw.

start h2o

title "H2O, cc-pVDZ basis, SCF optimized geometry"

geometry units au
H       0.0000000000   1.4140780900  -1.1031626600
H       0.0000000000  -1.4140780900  -1.1031626600
O       0.0000000000   0.0000000000  -0.0080100000
end

basis
H library cc-pVDZ
O library cc-pVDZ
end

scf
   thresh 1.0e-8
end

task scf

This extension may be omitted when calling the program. NWChem creates typically a whole array of output files that are documented in the User's Manual. A general log is displayed on the console, and may be saved in a file by redirecting the standard output:

nwchem sample > sample.log

where we assume that your input file is called sample.nw and you want to save the log to a file sample.log.

Note that this is just the basic syntax of the program call. In practise you will use a parallel environment to execute the program (see next section). In fact, executing NWChem by just typing the above line will run it in serial mode.

Parallel Runs

NWChem is inherently parallelized and designed to scale well on a multi-processor machine or a cluster. The underlying parallel system is MPI (Message Passing Interface) which is a commonly available standard that runs on many platforms. Consult our HowTo:mpi MPI help file and follow some of the links in there if you want to have more information about MPI. Even if you want to use only one processor for your NWChem run (which sometimes is the best solution, particularly for smaller computations), you have to submit the program to a parallel environment. On our clusters, the relevant command is mpirun:

mpirun -np 8 nwchem sample > sample.log

This will run your sample.nw input file on eight processors. Note that you are only allowed to run NWChem this way for small test systems! For any production jobs, you have to submit the task to the scheduler (see next section.).

Submitting (parallel) Gaussian jobs

If you want to run Gaussian on several processors on our machines, you have to include a line in your input file:

%nproc=8

where we assume that you want to use 8 processors (cores, threads).

It is mandatory to submit a Gaussian job script through our scheduling software (see our Scheduler Help File for details).

Here is a "bare bones" sample of a Gaussian submission script:

#! /bin/bash
#$ -S /bin/bash
#$ -q abaqus.q
#$ -l qname=abaqus.q
#$ -cwd
#$ -V
#$ -M hpcXXXX@localhost
#$ -m be
#$ -o STD.out
#$ -e STD.err
#$ -pe shm.pe 8
g09 < sample.g09 > sample.log
  • The first 6 lines of the script make sure the right shell is used, the program executes on the correct cluster, and all necessary setup is done.
  • An email address for notifications is specified in the #$ -M line. We suggest "hpcXXXX@localhost" (hpcXXXX stands for the username). Place a file .forward containing your actual email address into your home directory.
  • The -o and -e lines are used to tell the system where to write "standard output" and "standard error", i.e. the screen output.
  • The #$ -pe gaussian.pe 8 line specifies the number of processors the scheduler will allocate (8 in this example). It is crucial to choose the same number as specified in the %nrocs= line of the input file.

The script (let's call it g09.sh) is submitted by the qsub command:

qsub g09.sh

This must be done from the working directory, i.e. the directory that contains the input file and is supposed to contain the output. Also make sure that you have set up gaussian (use g09) before you submit a job.

Licensing

Gaussian is a licensed program. The license held by the Centre for Advanced Computing is limited to our computers at our main site. That means that any of our users can use the program on our machines (but nowhere else), whether they are located at Queen's or not.

We require users of Gaussian to sign a statement in which they state that they are informed about the terms of the license to be included in the Gaussian user group named "g98". Please fax the completed statement to (613) 533-2015 or scan/email to cac.admin@queensu.ca.

Where can I get more detailed information ?