Our current largest data set is SF30k. We should introduce SF100k and set the number of persons to achieve the size of 100TiB (in CSVs).