-
Notifications
You must be signed in to change notification settings - Fork 4
modules and compilation
Maxime Delorme edited this page Jul 15, 2025
·
1 revision
This page holds the information to compile the code on various supercomputers. All of these commands should be run on a reserved allocation on the machine (and not on a frontal node for instance).
module purge
module load arch/a100
module load cmake gcc/12.2.0 cuda/12.1.0 hdf5/1.12.0
cmake -DCMAKE_BUILD_TYPE=Release -DKokkos_ENABLE_CUDA=ON ..
#!/bin/bash
#SBATCH --job-name=[jobname]
#SBATCH -A [account]@a100
#SBATCH -C a100
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --gres=gpu:1
#SBATCH --cpus-per-task=24
#SBATCH --hint=nomultithread
#SBATCH --time=10:00:00
#SBATCH --output=log.out
#SBATCH --error=log.out
# modules to load for fv2d
module purge
module load arch/a100
module load cmake gcc/12.2.0 cuda/12.1.0 hdf5/1.12.0
# Echo of launched commands
set -x
srun ./fv2d [inifile]