HPC Software Modules (Python, PyTorch, Julia)

Overview

The Department of Statistics HPC provides centrally managed software via a shared /apps filesystem using Environment Modules.

This allows you to:

  • Load specific software versions on demand

  • Switch between versions easily

  • Use consistent, supported environments across all compute nodes

Available Software

You can see available modules with:

module avail

Example output:

/apps/modulefiles/Core:
  conda/miniforge3
  julia/1.11.3
  python/3.10.20
  python/3.11.15
  python/3.12.13
  pytorch/cu124-py310
  pytorch/cu124-py311
  aiml/py311

Basic Usage

Before using modules, initialise the module system:

source /etc/profile.d/modules.sh

Then:

module purge
module use /apps/modulefiles/Core
module load <module-name>

Examples

Python
module purge
module use /apps/modulefiles/Core
module load python/3.11.15

python3 --version
PyTorch (GPU)
module purge
module use /apps/modulefiles/Core
module load pytorch/cu124-py311

python -c "import torch; print(torch.cuda.is_available())"
Julia
module purge
module use /apps/modulefiles/Core
module load julia/1.11.3

julia --version
AI/ML Environment
module purge
module use /apps/modulefiles/Core
module load aiml/py311

python -c "import numpy, pandas, sklearn, torch; print('ok')"

Using Modules in Slurm Jobs

In Slurm jobs, the module system is not always automatically available.

You should initialise it explicitly.

Example:

#!/bin/bash
#SBATCH --job-name=test_python
#SBATCH --partition=standard-gpu
#SBATCH --clusters=srf_gpu_01
#SBATCH --gres=gpu:1
#SBATCH --cpus-per-task=2
#SBATCH --mem=4G
#SBATCH --time=00:05:00
#SBATCH --output=test_python_%j.out

set -euo pipefail

source /etc/profile.d/modules.sh

module purge
module use /apps/modulefiles/Core
module load python/3.10.20

echo "Python executable:"
which python3

echo "Python version:"
python3 --version

Common Issue: module: command not found

If you see:

module: command not found

initialise the module system first:

source /etc/profile.d/modules.sh

Conda / Miniforge

A shared Conda installation is available:

module load conda/miniforge3

Please see Storage for details on persistent vs temporary storage.

Persistent Conda Environment with Fast Package Cache

This is the recommended setup for most users.

Store the Conda environment persistently in /bitbucket/$USER/ while keeping the package cache on fast SSD-backed scratch storage.

Example:

module load conda/miniforge3

# Configure package cache on fast scratch
mkdir -p "$SCRATCH_FAST/conda-pkgs"

conda config --add pkgs_dirs "$SCRATCH_FAST/conda-pkgs"

# Create persistent environment
mamba create -y -p /bitbucket/$USER/conda-envs/myenv python=3.11

# Activate environment
source /apps/conda/miniforge3/bin/activate \
/bitbucket/$USER/conda-envs/myenv

Benefits:

  • Persistent environment

  • Faster package operations

  • Reduced load on home directories

Temporary Per-Job Environment

For temporary or rebuildable environments:

module load conda/miniforge3

mkdir -p "$SCRATCH_FAST/conda-envs"

mamba create -y \
-p "$SCRATCH_FAST/conda-envs/myenv" \
python=3.11

source /apps/conda/miniforge3/bin/activate \
"$SCRATCH_FAST/conda-envs/myenv"

Warning

/scratch storage is temporary and automatically wiped after 30 days.

Pixi / Rattler Cache

Pixi and Rattler perform many filesystem operations and may perform better with caches on fast scratch storage.

Example:

export RATTLER_CACHE_DIR="$SCRATCH_FAST/rattler-cache"
General Performance Guidance

For best performance:

  • Keep package caches on $SCRATCH_FAST

  • Avoid large Conda or Pixi caches in $HOME

  • Keep rebuildable environments on scratch storage where possible

  • Use persistent storage such as /bitbucket/$USER/ for environments you wish to keep long-term

Note

Home directories are currently HDD-backed storage, while /scratch/fast uses SSD-backed storage.

Performance is generally best when package caches are stored on $SCRATCH_FAST.

Best Practices

  • Always start with:

    module purge

  • Always load modules explicitly in scripts

  • Use shared modules for stable, supported environments

  • Use personal Conda environments for experimentation

Summary

The module system provides:

  • Consistent software across the cluster

  • Easy version switching

  • Preconfigured GPU and AI environments

If unsure, start with:

module avail