chriswasser
commented
5 days ago Additional information based on discussion after the RWTH Green500 presentation on 2024-07-16: Again, it was highlighted that there is a need to clarify the reporting of "10 average the power [...] measurements in the core phase". Proposal: Remove the need for reporting "10 average the power [...] measurements in the core phase" and just allow uploading of the raw, per measurement device CSVs in the associated data archive. Otherwise there are at least 3 different ways to aggregate unaligned / unsynchronized samples, which can skew the data in unintended ways. Having the raw data available allows any kind of postprocessing at a later stage if required.
Problem statement
In practical scenarios, a system is often covered by multiple measurement points. The document currently gives no guidance on how to arrive on a single average power value for the system, nor does it explain how to present the "10 average power / total energy measurements in the core phase.".
Solution
Write and include a text that explains how to do that properly.
The easiest way to get that right is to compute average core-phase power for each measurement domain, and then add those power values. However, this clashes with the requirement to include the time series in the report. I propose to keep the report flexible to allow multiple time series with different timestamps of different measurement domains.
Related issues
Examples - #104 Simplification of timing requirements - #108