↓ Skip to main content
The Login Node
The Login Node
  1. Talks & Workshops/

Introduction to Software Installation

HPC Software Installation Workshop Module Conda Spack EasyBuild Apptainer
Will Paik
Author
Will Paik
I optimize large-scale GPU clusters for AI/ML workloads. Outside of work, I build a mini-supercomputer from consumer hardware and document every step of it here.
Table of Contents

Date: July 17, 2026 Venue: Northeastern University, Boston, MA

Overview
#

A hands-on workshop for university researchers who need to get software running on a shared HPC cluster. This session walks through the main ways to install research software, from checking for an existing module to compiling from source, along with the pitfalls that trip people up on a shared, quota-limited system and working examples you can run yourself.

Topics Covered
#

  • Why software installation is different on a shared HPC cluster: no root access, shared filesystem, and tight quotas
  • Environment Modules: what module load actually does and why to check for existing software first
  • Conda: user-space package management, channels, and common HPC pitfalls (login node usage, home quota, activating environments in job scripts)
  • Spack: building a scientific software stack from source with control over compiler, MPI, and build variants
  • EasyBuild: installing from prebuilt easyconfigs on a fixed toolchain for reproducible builds
  • Installing from source: the standard build pattern (Autotools, CMake), linking against cluster libraries, and writing personal modulefiles
  • Apptainer: running and building containers on a shared cluster without root access
  • Choosing the right tool for the job

Materials
#