Biggest reason Data Center managers won't dump Intel for ARM Server chips
ARMH can't guarantee them that they can keep up with Intel's advances over time. Managers aren't going to make such a humongous change for chips that might not even be competitive in a year.
Data Center manager to salesrep: What's your R & D buget compared to Intel's? Gulp
Datacenter mananger to salesrep: Intel controls their own destiny since they manufacture their own chips. Who manufactures yours? Is it here in the U.S.? Gulp.
I read in an article on SA that Microsoft is working on porting over Windows to ARM, What, that's not true? :)
There is a nice, lightweight copy of Windows for ARM that can probably work with some minor tweaking. Let's call it a weekend project for the boys at Cambridge. It's called Windows NT. They can at least change the splash screen and fool a few journalists. That should be enough to get ARMH to over $60.
First there needs some really good reasons for a datacenter manager to even consider switching from Intel architecture to ARM. Switching is a major undertaking and requires significant time, labor and expense. The ARM camp has made a lot of made bold claims about supposed energy efficiency & performance superiority but the results aren't there. Even inside investors are abandoning the ARM server dream. What the investors saw is what everyone else sees - there's no payback to justify the risk.
And you are correct that the decision on architecture isn't just a for a single product generation, it's a strategic decision that forces decision makers to consider the vendor's experience, reputation, longevity, & stability.