Welcome to the GitHub profile of Fraunhofer IISB! Here you can find our open-source contributions and projects.
The Fraunhofer Institute for Integrated Systems and Device Technology IISB conducts applied research and development in power electronics, semiconductor devices, and intelligent electronic systems. Key areas include energy-efficient power modules, wide-bandgap semiconductors (SiC/GaN), battery systems, AI-supported simulation & optimization, thermal/electro-thermal modeling, and reliable electronics for automotive, industrial, aerospace, and energy applications. We bridge the gap between fundamental research and industrial application — developing technologies that are efficient, robust, sustainable, and ready for the future.
Are you passionate about cutting-edge technologies in power electronics, semiconductors, AI-driven simulation or next-generation energy systems? Join our team at Fraunhofer IISB in Erlangen — we offer exciting positions for scientists, engineers, PhD students, and interns.
- Website: https://www.iisb.fraunhofer.de/
- LinkedIn: Fraunhofer IISB
- jNO (jax Neural Operators) is a JAX-native library for neural operators and foundation models with unified support for both data-driven and physics-informed training. Its core design is a tracing system in which domains, model calls, residuals, supervised losses, and diagnostics are written in one symbolic language and compiled into one optimization pipeline. This allows users to move between operator regression, mesh-aware residual evaluation, and PDE-constrained training without restructuring the surrounding code. jNO also supports multi-model compositions, fine-grained control at parameter level (model, optimizer, and learning rate), hyperparameter tuning, and JAX-native workflows for translated PDE foundation-model families.
- foundax is a library for foundation model and neural operator architectures.
- Hard-constraining Neumann - Hard-constraining Neumann boundary conditions in physics-informed neural networks via Fourier feature embeddings
- CleanQRL - Clean and easy to understand implementations of many Quantum Reinforcement Learning agents as well as their classical analouges.
- ParaPowerPython - Python adaptation of the (open-source) thermal solver ParaPower (originally written in MATLAB) for rapid evaluation of 3D power module thermal behavior using a lumped thermal resistance network.