This hands-on workshop introduces participants to containerization in High-Performance Computing (HPC), focusing on Apptainer (formerly Singularity) as the primary runtime for reproducible scientific workloads on clusters such as amplitUDE.
Participants will explore the container ecosystem, understand how Docker, Docker Hub, and NVIDIA GPU containers integrate into HPC workflows, and gain practical experience building and running containers on CPU and GPU nodes.
By the end of this workshop, participants will be able to:
-
Explain how containerization enhances reproducibility, portability, and sustainability in scientific research.
-
Compare Docker, Apptainer, and virtual machines in terms of performance, privileges, and HPC compatibility.
-
Use Docker Hub and other container registries as sources for reproducible environments.
-
Build and run custom containers with Apptainer, including NVIDIA GPU support.
-
Execute containerized applications within Slurm job scripts on amplitUDE.
Prerequisites:
- Basic knowledge of the Unix Shell, e.g., from the carpentry course.
- Access to amplitUDE or any HPC system with Apptainer/Singularity available.
- Bring your own laptop (preferably Linux).
- begleitende Lehrperson: Dominik Brands
- begleitende Lehrperson: Christoph Jokubonis
- verantwortliche Lehrperson: Junaid Mir