HPC/Applications/lammps: Difference between revisions
< HPC | Applications
Jump to navigation
Jump to search
m (→Benchmark) |
m (→Benchmark) |
||
| Line 6: | Line 6: | ||
{| class="wikitable" cellpadding="5" style="text-align:center; margin: 1em auto 1em 1em;" | {| class="wikitable" cellpadding="5" style="text-align:center; margin: 1em auto 1em 1em;" | ||
|- style="background:#eee;" | |- style="background:#eee;" | ||
! Job | ! Job tag || Node type !! Interconnect !! Additional OpenMPI options !! Relative speed<br>(1000 steps/3 hours) !! Notes | ||
|- | |- | ||
| gen1 || gen1 || IB || (none) || 36 | | gen1 || gen1 || IB || (none) || 36 | ||
| Line 14: | Line 14: | ||
--> | --> | ||
|- | |- | ||
| gen1srqpin || gen1 || IB || -mca btl_openib_use_srq 1<br>-mca mpi_paffinity_alone 1 || 39 | | gen1srqpin || gen1 || IB || [http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/9333.html -mca btl_openib_use_srq 1<br>-mca mpi_paffinity_alone 1] || 39 | ||
<!-- | <!-- | ||
|- style="background:#e8e8e8;" | |- style="background:#e8e8e8;" | ||
Revision as of 21:56, January 23, 2012
Benchmark
Using a sample workload from Sanket ("run9"), I tested various OpenMPI options on both node types.
LAMMPS performs best on gen2 nodes without extra options, and pretty well on gen1 nodes over ethernet(!).
| Job tag | Node type | Interconnect | Additional OpenMPI options | Relative speed (1000 steps/3 hours) |
Notes |
|---|---|---|---|---|---|
| gen1 | gen1 | IB | (none) | 36 | |
| gen1srqpin | gen1 | IB | -mca btl_openib_use_srq 1 -mca mpi_paffinity_alone 1 |
39 | |
| gen1eth | gen1 | Ethernet | -mca btl self,tcp | 44 | fastest for gen1 |
| gen2eth | gen2 | Ethernet | -mca btl self,tcp | 49 | |
| gen2srq | gen2 | IB | -mca btl_openib_use_srq 1 | 59 | |
| gen2 | gen2 | IB | (none) | 59 | fastest for gen2 |
Sample job file gen1
#!/bin/bash
#PBS -l nodes=20:ppn=8:gen1
#PBS -l walltime=3:00:00
#PBS -N <jobname>
#PBS -A <account>
#
#PBS -o job.out
#PBS -e job.err
#PBS -m ea
# change into the directory where qsub will be executed
cd $PBS_O_WORKDIR
mpirun -machinefile $PBS_NODEFILE \
-np $(wc -l < $PBS_NODEFILE) \
-mca btl self,tcp \
lmp_openmpi < lammps.in > lammps.out 2> lammps.err
Sample job file gen2
#!/bin/bash
#PBS -l nodes=20:ppn=8:gen2
#PBS -l walltime=3:00:00
#PBS -N <jobname>
#PBS -A <account>
#
#PBS -o job.out
#PBS -e job.err
#PBS -m ea
# change into the directory where qsub will be executed
cd $PBS_O_WORKDIR
mpirun -machinefile $PBS_NODEFILE \
-np $(wc -l < $PBS_NODEFILE) \
lmp_openmpi < lammps.in > lammps.out 2> lammps.err