Sustainability in Research Software for HPC
by
Building 20.20, Room 167
KIT Campus South

Maximal performance! This used to be the goal for research software for HPC, but sustainability is getting more and more traction and growing importance. This course targets two aspects contributing to sustainable research software for HPC clusters:
Firstly, energy and carbon optimization in the usage is outlined to raise awareness and make the participants ready for future HPC systems with energy-budget-based resource grants or efficiency rewarding queuing systems as pioneered by Japans #1 supercomputer Fugaku. Secondly, sustainability in the development process of HPC research software is addressed, where productivity, code quality and carbon efficiency often go hand in hand. The course is specifically targeted to researchers using and developing HPC research software, in all career stages, irrespective of their usage type (e.g. physical simulations, data analysis, AI).
It is intended as an interactive course, allowing plenty of time for discussion and exchange on this topic.
The course will take place on-site (in-person session). It will be conducted in English.