1 Install NAMD2.13 at HPE SGI 8600 using intel Omni-Path networks

Click Start Over at the left bottom to start Back to Content

1.1 Run NAMD

MPI version was compiled with TCL or FFTW, since Charm++ MPI network layer performs better than the verbs layer, and Intel Omni-Path networks are incompatible with the pre-built ibverbs NAMD binaries.

NAMD MPI binaries may be launched with mpiexec:

  • setenv MPI_DSM_VERBOSE 1
  • setenv MPI_MEMMAP_OFF true
  • mpiexec_mpt -np $npr namd2 conf

1.2 Compile NAMD

The following is the building process.

  • tar xzf NAMD_2.13_Source.tar.gz
  • cd NAMD_2.13_Source
  • tar xf charm-6.8.2.tar
  • Build and test the Charm++/Converse library (MPI version):
    • cd charm-6.8.2
    • env MPICXX=mpicxx ./build charm++ mpi-linux-x86_64 --with-production
    • cd mpi-linux-x86_64/tests/charm++/megatest
    • make pgm
    • mpiexec -n 4 ./pgm (run as any other MPI program on your cluster)
    • cd ../../../../..
  • Download and install TCL and FFTW libraries: (cd to NAMD_2.13_Source if you’re not already there)
  • Set up build directory and compile:
    • MPI version: ./config Linux-x86_64-g++ --charm-arch mpi-linux-x86_64
    • cd Linux-x86_64-g++
    • gmake -j4 (which should run faster than make)