Hi Arisa,
I remember Wailnn brought up this concern, but I am not able to find that email chain-- maybe Timo or Wailnn can find it. I think we agreed that for any RGA by week there are at least 10 wafers represented. For 509.04L there are three weeks of RGA so each week needs to have 10 wafer represented. However, within a given RGAweek, there are multiple configurations/sbrs and it is difficult to have 10 for each configuration/sbr. So we agreed on 4 wafers represented for each configuration/sbr (4 is what I remember, but Timo/Wailnn can confirm). THat way we don't have all TI/MPT with one wafer and LSI/NHK with a different wafer causing wrong conclusion when we analyze the results of the RGA.