This reminds me of the kind of "IDC deceit" that goes on with Mac vs PC marketshare statistics. (if you subtract the millions of windows drone-boxes its a different picture)
I'd bet you a sweet nickel that a huge number of those windows servers are in small office situations. If you took the stats for servers in data centers, and/or the ones doing heavy lifting I suspect it would still be very heavily unix/linux based.
For example, in the Top 500 supercomputers, a full 0.20 % are Windows based.
TWO TENTHS OF ONE PERCENT.
Windows servers ...my sweet arse!
Last edited by Script Kiddie; Feb 24th, 2006 at 07:05 AM.
FWIW, they are comparing revenue - not actual numbers of how many are sold.
Considering that in order to run a Microsoft server you need to buy the Server OS and then buy licenses for every user that will use it, compared to a lower cost Linux solution (but they usually come with pretty decent support packages if paid for).
This is kind of like saying that since I sold 3 Hummers at $100k, and only 9 Kias at $10k the market share of hummers is higher, when in reality the Kias outnumber the Hummers 3:1.
So in summary, I'm of the opinion the conclusion they take from the benchmark they are using is complete BS.