onsite
Research Software Engineer - Coupled Climate Models on GPU Based HPC Systems - Deutsches Klima rechenzentrum GmbH
Software Engineer
Lead the development and optimization of coupled climate models on GPU‑based high‑performance computing platforms, driving scientific discovery through advanced software engineering and performance tuning.
About the role
Key Responsibilities
- Design, implement, and maintain scalable coupled climate model components for GPU‑accelerated HPC environments.
- Collaborate with climate scientists to translate research requirements into efficient, reproducible code.
- Profile and optimize code performance, ensuring optimal utilization of GPU resources and memory bandwidth.
- Integrate and validate model outputs against observational datasets and benchmark experiments.
- Document software architecture, performance results, and best practices for future development.
Requirements
- Strong programming skills in Python and C++ with experience in GPU programming (CUDA, OpenACC, or similar).
- Proven track record of working on large‑scale scientific simulations or climate modeling projects.
- Deep understanding of HPC concepts, parallel computing, and performance analysis tools.
- Excellent problem‑solving abilities and a collaborative mindset.
- Fluency in English; German language skills are a plus.