For requests and questions with regard to the HPC Performance Lab contact firstname.lastname@example.org.
Other support channels:
- Once per month the RRZE organizes the HPC-Cafe where you can meet us in person.
- For issues concerning RRZE HPC systems contact email@example.com
- For KONWIHR questions and project applications contact firstname.lastname@example.org
We provide the following services:
- General HPC user support, e.g., single core optimization, MPI and hybrid parallelization, compiler issues, or I/O improvements
- Node-level performance engineering (PE) for CPUs and GPUs, including performance analysis, analytic performance modeling, and optimization actions (e.g., appropriate blocking techniques, parallelization strategy, or efficient data layouts)
- Microbenchmarking and architectural analysis. The RRZE HPC Testcluster provides access to a wide range of CPUs and GPUs.
- Algorithmic expertise is provided for numerical methods for simulations and software-libraries in the field of iterative and especially sparse solvers like multigrid methods and lattice Boltzmann solvers.
- Trainings and Tutorials about programming for HPC systems and Node-level Performance Engineering