Job scripts (Tegner)

This is an example of a job script for a MPI program. For other programs, you can find examples in the software page Software.

#!/bin/bash -l
# The -l above is required to get the full environment with modules

# Set the allocation to be charged for this job
# not required if you have set a default allocation
#SBATCH -A <201X-X-XX>

# The name of the script is myjob
#SBATCH -J myjob

# Only 1 hour wall-clock time will be given to this job
#SBATCH -t 1:00:00

# Number of nodes
#SBATCH --nodes=2
# Number of MPI processes per node
#SBATCH --ntasks-per-node=24

# Type of GPU for CUDA code
#SBATCH --gres=gpu:K80:2

# Load the Intel MPI module
module add intelmpi/17.0.1

# Run the executable named myexe with MPI-rank of 48
# and write the output into my_output_file
mpirun -n 48 ./myexe > my_output_file 2>&1

SLURM Flags for specific nodes

There are different types of compute nodes on tegner, see Hardware. You can specify the nodes using the SLURM flags shown in the following table.

Computer nodes

Details

SLURM flag

2 TB Fat nodes

total 5 nodes 4x12 core E7-8857v2 (48 cores per node) 2 TB RAM per node 2x Nvidia Quadro K420 GPU

--mem=2000000

1 TB Fat nodes

total 5 nodes 4x12 core E7-8857v2 (48 cores per node) 1 TB RAM per node 2x Nvidia Quadro K420 GPU

--mem=1000000

Thin nodes with K420 GPU

total 46 nodes 2x12 core E5-2690v3 (24 cores per node) 512 GB RAM per node Nvidia Quadro K420 GPU

--gres=gpu:K420:1

Thin nodes with K80 GPU

total 9 nodes 2x12 core E5-2690v3 (24 cores per node) 512 GB RAM per node Nvidia Quadro K80 GPU

--gres=gpu:K80:2

NBIS node

total 1 node 14x12 core (168 cores per node) 4 TB RAM per node

-w t04n03 --mem=4000000