It would be useful to save more information from the log files to extract the bottlenecks of individual runs.
In a first step one should trim and save the log files in the generate directories.
In a next step one could think about adding a -v flag to mdbenchmark analyze which would show tables comparing the individual contributions to the runtime.
It would be useful to save more information from the log files to extract the bottlenecks of individual runs.
In a first step one should trim and save the log files in the generate directories.
In a next step one could think about adding a
-vflag tomdbenchmark analyzewhich would show tables comparing the individual contributions to the runtime.