Created by Frank Schluenzen, last modified on Jan 13, 2022 09:49

Summary

  • Source: http://www.ks.uiuc.edu/Research/namd/
  • License:  see http://www.ks.uiuc.edu/Research/namd/license.html
  • Path:  /opt/namd
  • Documentation: http://www.ks.uiuc.edu/Research/namd/2.10/ug/
  • Citation: the user agrees that any reports or published results obtained with the Software will acknowledge its use by the appropriate citation as follows: "NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign." Any published work which utilizes NAMD shall include the following reference: "James C. Phillips, Rosemary Braun, Wei Wang, James Gumbart, Emad Tajkhorshid, Elizabeth Villa, Christophe Chipot, Robert D. Skeel, Laxmikant Kale, and Klaus Schulten. Scalable molecular dynamics with NAMD. Journal of Computational Chemistry, 26:1781-1802, 2005.". Electronic documents will include a direct link to the official NAMD page at http://www.ks.uiuc.edu/Research/namd/

NAMD s a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. 

Based on Charm++ parallel objects, NAMD scales to hundreds of cores for typical simulations and beyond 500,000 cores for the largest simulations. NAMD uses the popular molecular graphics program VMD for simulation setup and trajectory analysis, but is also file-compatible with AMBER, CHARMM, and X-PLOR. NAMD is distributed free of charge with source code. You can build NAMD yourself or download binaries for a wide variety of platforms

Using  namd

namd2 comes in a variety of different versions tailored for specific hardware environments. On the maxwell-cluster the ibverbs or ibverbs-cuda versions are probably most efficient. To initialize the specific environment use the module command:

[elx]% module avail
[elx]% module load namd/ibverbs           # Linux-x86_64-ibverbs (InfiniBand via OpenFabrics OFED, no MPI needed)
[elx]% module load namd/ibverbs-smp-CUDA  # Linux-x86_64-ibverbs-smp (InfiniBand plus shared memory, no MPI needed)
[elx]% module load namd/multicore         # Linux-x86_64-multicore (64-bit Intel/AMD single node)
[elx]% module load namd/multicore-CUDA    # Linux-x86_64-multicore-CUDA (NVIDIA CUDA acceleration)
[elx]% module load namd/multicore-MIC     # Linux-x86_64-multicore-MIC (Intel Xeon Phi coprocessor acceleration)
[elx]% module load namd/tcp               # Linux-x86_64-TCP (multi-core, multi-host using ethernet (TCP) for process communication)