Computer Applications in Physics: Parallelization and High-Performance Computing (UHH, WiSe 2024/25)

This repository contains exercises and example code from the course
Computer Applications in Physics: Parallelization and High-Performance Computing
by Martin Stieben (PHYSnet) and Mohammed El-Nabulsi (ILP), University of Hamburg.

📌 Course Topics

• Shared-memory parallelization with multithreading (Numba)
• Distributed-memory parallelization with MPI
• GPU acceleration for computational routines
• Performance benchmarking and scaling (weak/strong)
• Data storage and compression with HDF5
• Practical use of the PHYSnet compute cluster

📂 Repository Structure

• exercises/ – Python notebooks and scripts for each exercise sheet
• data/ – sample datasets and HDF5 storage examples
• requirements.txt – pip dependencies
• environment.yml – Conda environment setup

▶️ Getting Started

1. Clone the repo:

   git clone https://github.com/<your-username>/hpc-exercises-uhh-2024-25.git
   cd hpc-exercises-uhh-2024-25

2. Create environment (Conda, recommended):

   conda env create -f environment.yml
   conda activate hpc-uhh

3. Run JupyterLab or your preferred editor:

   jupyter lab

⚙️ Dependencies

• Python >= 3.9
• numpy, scipy, matplotlib
• numba (CPU multithreading)
• mpi4py (MPI-based parallelization)
• h5py (HDF5 I/O)
• jupyterlab

See requirements.txt or environment.yml for install details.

📜 License

MIT License – free to use and adapt.