[SC23] metrics litterature
In order to find relevant metrics for Melissa, this issue presents various performance/efficiency metrics that were used in SC22 papers.
Amongst them, 13 papers were selected from their title. After rapid examination, 5 were retained:
This paper presents an in-situ visualization frameworks that has strong similarities with Melissa:
Melissa is an in transit service for per- forming large-scale sensitivity analysis that bears similarities to SERVIZ in its design. In particular, both follow an explicit client-server model and perform an MxN data re-distribution using a two-level approach. However, the primary motivation for Melissa is to avoid or reduce the number of parallel file system operations — in contrast, SERVIZ aims to provide cost savings over existing in situ analysis techniques.
Efficiency indicators presented there may be worth considering as a basis to derive our own.
In addition, detailed break-up of the execution time (Fig. 4, 5 and 8) could easily be adapted to our context.