Introduction

This repository has all the scripts and inputs necessary to run a toy model of the ocean using MOM6. A machine-learning model is embedded into the numerical solver (the momentum component of the Navier-Stokes equations).

Quick Instructions

1. Navigate to the workshop scratch space on the Lustre filesystem

   export SCRATCH_ROOT=/lustre/orion/gen150/world-shared/smartsim-2023/workshop_scratch/$USER
   mkdir -p $SCRATCH_ROOT
   cd $SCRATCH_ROOT
   

2. Clone down the repository

   git clone https://github.com/CrayLabs/OLCF_SmartSim2023.git
   

3. Get a Slurm allocation

   salloc -N 1 -t 01:00:00 --reservation=smartsim -A YOUR_ALLOCATION_NAME_HERE
   

4. Add the modulefile for SmartSim

   module use $SCRATCH_ROOT/OLCF_SmartSim2023/modulefiles
   

5. Load the modulefile

   module load py-smartsim-0.5.0-gcc-11.2.0
   

6. Run the MOM6 example

   cd $SCRATCH_ROOT/OLCF_SmartSim2023/smartsim_drivers/double_gyre
   ulimit -s unlimited # Necessary for MOM6
   python call_MOM6.py