Working Towards FAIRness in Performance Data

Making sense of the performance data generated by computational fluid dynamics (CFD) experiments is challenging, both due to the high volume of data and the high variability of possible inputs. We present preliminary work on a system to store sufficient metadata alongside the performance data generated by the Score-P measurement tool to make these results more Findable, Accessible, Interoperable, and Reusable (FAIR). This system can be deployed in both continuous integration and production runs of CFD codes to allow developers better insight into potential causes of performance variations.