Last week, I attended the HLRS Results & Review Workshop. Think of it as a kind of HLRS user group meeting. Given my background as computational scientist, numerical astrophysics more specifically, it was very intersting to see the full depth and breadth of scientific projects running on HLRS systems. Obviously, code performance and efficient use of allocated resources is also on everybody's mind during such an event.

The organizers kindly allowed me to present the POP project to HLRS users. As HLRS staff, I am in regular contact with some of our users, but seldom have the chance to adress most if not all of them at the same time. During the following poster session, I was engaged in many interesting discussions on the parallel performance of our user's codes. Incidentally, I met a participant of one of our code optimisation courses last year. He told me, that he had improved the performance of his code by a factor of ten, and he was confident he now could run the largest simulation of this kind.

I hope some HLRS users will contact POP to request performance assessment services and look forward to work with them.

José Gracia (HLRS)