NAMD

Overview

NAMD is a scalable molecular dynamics package written at the University of Illinois at Urbana Champaign which is particularly suited to biological applications.

NAMD is one of a new generation of molecular dynamics codes which are designed with a distributed data model to improve scalability on large numbers of processors. NAMD scales well for large problems up to 256 or 500 processors on the CSAR Origin systems.

NAMD uses Protein Data Bank files as the standard files for describing the configuration of atoms involved in the simulation, but it also provides compatibility with Amber parm and coordinate files, as well as Gromacs files. Unless specified otherwise, NAMD uses the CHARMM force field, but it is also able to implement the Amber force field.

The latest version of NAMD (2.6b1; release date 30 July 2005) is now available on Newton. NAMD 2.6b1 has several advantages over NAMD 2.5, including:

• Adaptive biasing force free energy calculations.
• Tcl-based boundary potentials
• Reduced memory usage for unusual simulations
• Support for OPLS forces field

Restrictions on Use

There are no restrictions on using the NAMD software.

Set Up Procedure

In order to run NAMD you should load the namd module by typing module load namd

Running the Code

Namd is an MPI application and the executable is called namd2. To launch a 64 processor application you would add the following to your batch script

• On the Origins mpirun -np 64 namd2
• On the Altix set the number of processors to be 64 using the -n option to bsub and then add mpirun -np 64 /usr/local/apps/supported/namd/2.6b1/NAMD_2.6b1_Linux-ia64-MPT/namd2

It is also important that you have set the environment variable MPI_DSM_DISTRIBUTE. This can be done with

export MPI_DSM_DISTRIBUTE=yes (for bash-like shells)

setenv MPI_DSM_DISTRIBUTE yes (for csh-like shells)

Further Information

For further information please look at the NAMD Web pages.