Dave's preliminary comment's on offline computing spreadsheet numbers
Revision as of 08:53, 27 April 2012 by Davidl (Created page with "''Discussed at Apr. 27, 2012 Software Review Planning Meeting'' # Consider replacing "CPU" with "Core" # CPU-s/event i...")
Discussed at Apr. 27, 2012 Software Review Planning Meeting
- Consider replacing "CPU" with "Core"
- CPU-s/event is probably closer to <=100ms now. This can be checked easily.
- pass 0 CPU is based on spreading job out over full year. Should this be over 2-3 weeks?
- Does MC time per event generation include mcsmear? Should this be listed separately?
- MC event size should be bigger than raw data. Maybe by factor of 2 since it holds smeared and unsmeared values. (former could be dropped if needed)