HOOMD-blue is a general purpose particle simulation toolkit. It performs hard particle Monte Carlo simulations of a variety of shape classes, and molecular dynamics simulations of particles with a range of pair, bond, angle, and other potentials. HOOMD-blue runs fast on NVIDIA GPUs, and can scale across thousands of nodes.

To use Hoomd, you need to load the appropriate modules in your Grid Engine submit file.

#$/bin/bash
#
#$ -cwd
#$ -j y
#$ -S /bin/bash

. /etc/profile.d/modules.sh

module load python/3.7.2 cuda91/toolkit openmpi/3.0.0 intel-tbb-oss/intel64/2019_20191006oss

Here is a simple example, save this file to a file called "run.py"

import hoomd
from hoomd import md
hoomd.context.initialize()

# Create a 10x10x10 simple cubic lattice of particles with type name A
hoomd.init.create_lattice(unitcell=hoomd.lattice.sc(a=2.0, type_name='A'), n=10)

# Specify Lennard-Jones interactions between particle pairs
nl = md.nlist.cell()
lj = md.pair.lj(r_cut=3.0, nlist=nl)
lj.pair_coeff.set('A', 'A', epsilon=1.0, sigma=1.0)

# Integrate at constant temperature
md.integrate.mode_standard(dt=0.005)
hoomd.md.integrate.langevin(group=hoomd.group.all(), kT=1.2, seed=4)

# Run for 10,000 time steps
hoomd.run(10e3)

Now lets run "run.py" in GPU mode by requesting one of the GPU nodes

#$/bin/bash
#
#$ -cwd
#$ -j y
#$ -q gpu.q
#$ -S /bin/bash
. /etc/profile.d/modules.sh
module load python/3.7.2 cuda91/toolkit openmpi/3.0.0 intel-tbb-oss/intel64/2019_20191006oss
python run.py

You can safely ignore the warning messages about mca_base_component_repository_open in your job output file because we build our own OpenMPI statically.

-- AdminUser - 27 Nov 2019