As HPC platforms grow in scale and complexity, understanding the performance problems of applications becomes essential to run efficiently on these machines. Many bottlenecks however, only manifest under real production conditions, meaning that smaller-scale tests are at the risk of optimizing things, that will not be problematic at production scale, while real bottlenecks might never arise.

In this webinar we will showcase how POP tools can be used to obtain insight into the behavior of the application, even at production scale.
We present in-depth analysis of Vlasiator, a plasma simulation code that models the collisionless plasma environment surrounding the earth.
Enabled by the insight we were able to provide a second level service fixing a performance bug in the field solvers, which results in measurable improvements at production scale.