Does AI not realize how useless and flawed the data from CIRP is? They are essentially two friends with a blog, who only survey 500 people to get their "data". You don't need to be a statistics major to know that that sample size is far too small to be salient, even if they did sample it properly (which they don't).
Their data is completely worthless, and shame on publications that report on their junk "data" without doing any due diligence.
I guess that answers my question… how did they account for institutional buyers buying crate loads of them for embedded uses and data centres.
Simple they didn’t.
This is correct. The CIRP survey is worthless.
Exhibit A: in the early years of the Mac mini, it was heavily deployed in Las Vegas casinos in conjunction with cameras for video surveillance of gaming tables.
Exhibit B: also from the early days, the Mac mini was installed in colocation facilities as servers. Some of those companies still exist. Unless happened to interview one of those IT directors, those units would not be part of their dataset.
From the beginning the Mac mini was frequently marketed as a SOHO device, for printer serving, data storage, etc. Again it is unlikely CIPR's sample of interviewees properly captures this market.
Comments
Exhibit A: in the early years of the Mac mini, it was heavily deployed in Las Vegas casinos in conjunction with cameras for video surveillance of gaming tables.
Exhibit B: also from the early days, the Mac mini was installed in colocation facilities as servers. Some of those companies still exist. Unless happened to interview one of those IT directors, those units would not be part of their dataset.
From the beginning the Mac mini was frequently marketed as a SOHO device, for printer serving, data storage, etc. Again it is unlikely CIPR's sample of interviewees properly captures this market.