Running and Verifying LAMMPS
Procedure
- Use PuTTY to log in to the server as the root user.
- Run the following command to create the working directory:
mkdir -p path/to/CASE
- Perform the following commands to switch to the working directory and copy the computing instance and binary files to the working directory.
cd /path/to/CASE cp /path/to/LAMMPS/lammps-stable_5Jun2019/bench/in.lj ./ cp /path/to/LAMMPS/lammps-stable_5Jun2019/src/lmp_g++_openmpi ./
- Run the following command:
mpirun --allow-run-as-root -np 96 --mca btl ^openib ./lmp_g++_openmpi -in in.lj
Check the value (in the unit of timesteps/s) of Performance in the log.lammps log file. A higher value indicates better performance.
The following is an example of the test result.
Performance: 593019.799 tau/day, 1372.731 timesteps/s 98.1% CPU use with 96 MPI tasks x no OpenMP threads MPI task timing breakdown: Section | min time | avg time | max time |%varavg| %total --------------------------------------------------------------- Pair | 0.028215 | 0.040367 | 0.051175 | 3.4 | 55.41 Neigh | 0.0036562 | 0.0049894 | 0.0061901 | 1.2 | 6.85 Comm | 0.013968 | 0.026072 | 0.039602 | 4.8 | 35.79 Output | 7.905e-05 | 0.000128 | 0.00023195 | 0.0 | 0.18 Modify | 0.00056847 | 0.00080538 | 0.00099514 | 0.0 | 1.11 Other | | 0.0004857 | | | 0.67
Parent topic: LAMMPS 5 Jun 2019 Porting Guide (CentOS 7.6)