We adapted the example prepared by the Oceananigans' developer to run on Gadi. The test is done on both normalbw and gpuvolta nodes.  Follow the instructions below to run the shallow water model simulation on both CPUs and a single GPU. 

Use the Shared Environment

Here is a minimal example to use the shared environment in a shell. To try it in a jupyter notebook, follow the example in the next section.

$ module use /g/data/dk92/apps/Modules/modulefiles/
$ module load NCI-data-analysis/2022.06
$ julia

julia> sdir="/g/data/dk92/notebooks/examples-julia/Oceananigans/"
julia> pushfirst!(DEPOT_PATH,sdir*"SharedEnv")
julia> pushfirst!(DEPOT_PATH,mkpath(joinpath(ENV["TMPDIR"],"env")))
julia> using Oceananigans

Run the tested notebook

  1. Go to ARE site: are.nci.org.au
  2. Fill out the JupyterLab request form
    1. Walltime (hours): 4
    2. Queue: gpuvolta
    3. Compute Size: 1gpu
    4. Project:  <xy01>
    5. Storage: gdata/dk92+scratch/<xy01>
  3. Click Advanced options and fill in the following fields
    1. Module directories: /g/data/dk92/apps/Modules/modulefiles/
    2. Modules: NCI-data-analysis/2022.06 cuda/11.4.1
    3. Jobfs size: 100GB  
  4. Launch the session to run the tested notebook

Note on the tested notebook

  • copy the tested notebook from "/g/data/dk92/notebooks/examples-julia/Oceananigans/shallow_water_Bickley_jet.ipynb" and/or "/g/data/dk92/notebooks/examples-julia/Oceananigans/cpu_vs_gpu" to your own working directory
  • the notebook shallow_water_Bickley_jet.ipynb contains an example to run a simulation.
  • the notebook cpu_vs_gpu.ipynb contains an example of comparison between simulaitons that run on CPU and GPU.
  • if necessary, modify DEPOT_PATH in the notebook to use another environment. The default one points to "/g/data/dk92/notebooks/examples-julia/Oceananigans/SharedEnv"



  • No labels