CPU ratios for nightly tests

I'm posting plots of the results of several of the nightly test jobs from the past 6 days, where the horizontal axis is the 500 events processed nightly (0-499 is Saturday, 500-999 is Sunday, ...etc... , 2500-2999 is Thursday). I then take the ratio event-by-event for the 4 types of jobs (unoptimized vs. optimized and 32-bit vs. 64-bit).

Currently, only 5 real data (DAQ) jobs are done with the 4 job types.
  • daq_sl302.ittf/.../year_2014/AuAu200_production_mid_2014 :

     
  • daq_sl302.ittf/.../year_2015/production_pAu200_2015 :
    (NB: 1000 events each night instead of 500)

     
  • daq_sl302.stica/.../year_2016/AuAu200_production_2016 :

     
  • daq_sl302.stica/.../year_2017/AuAu54_production_2017 ;

     
  • daq_sl302.stica/.../year_2018/27GeV_production_2018 :

     
That's it.

-Gene