>While Intel and MS will make money hand over first for years to come, it does appear to be the end of the consumer market for these two companies. Their focus will be business class computers, workstations, and servers.
the server market, for all of my career, really, has been following the consumer market. The stuff you see in a server is basically upmarket desktop kit; the primary difference being the addition of ecc in various places.
If ARM beats X86 on the desktop, within a short number of years, they will beat x86 on the server, too.
Now, arm is a /long way/ from winning the desktop market, and I'm not at all sure they will. I'm just saying... /if/ x86 loses the consumer market, they will also lose the cost-sensitive server market.
x86 only owns the server market because the r&d and economies of scale paid for by the consumer market.
Personally, I find that... unlikely. Of course, that's a market I know nothing about, so I could very well be wrong. It doesn't really change my argument, though. s/desktop/consumer hardware/g and you get the same meaning. It's the mass market that makes things cheap, and if you can cram enough power in to a tablet CPU that it can replace the desktop, you can bet that us server folks will start using those CPUs (maybe modified with ecc support in the ram/cache) for our own applications
The real difference for us server folk is that tablet CPUs are always going to be behind the curve in terms of raw power. Likely, that will result in server boards accelerating further in to the 'massively multicore' direction they are going now. Which is okay with me; 1024 chickens, for what I do, actually work better than four strong oxen.
You really see offices full of workers staring at tablet PCs, trains & planes full of road warriors bashing out long reports on touchscreens?
The home market may primarily shift to tablets; I'm not convinced by that and I think there's still a problem with the use case for content creators on tablets which includes an awful lot of school children. But the market I absolutely can't see abandoning PCs as we see them now and moving to handheld touchscreens is the office market and that's far from insignificant.
Tablets cost considerably less than desktop computers. There is nothing preventing you from being able to plug a second monitor and a mouse into a tablet and use it exactly the same way that you use a laptop and a desktop today.
At the moment, they're significantly more expensive and less compatible. They're also, by virtue of design, far easier to steal / lose and we've had quite enough (UK) stories about laptops being left on trains by mistake or stolen from homes and critical data being leaked as a consequence. Also, I can't see many people wanting a 10" desktop monitor or a 17", let alone 22", tablet.
I don't dispute it could happen, but I'd be surprised.
the server market, for all of my career, really, has been following the consumer market. The stuff you see in a server is basically upmarket desktop kit; the primary difference being the addition of ecc in various places.
If ARM beats X86 on the desktop, within a short number of years, they will beat x86 on the server, too.
Now, arm is a /long way/ from winning the desktop market, and I'm not at all sure they will. I'm just saying... /if/ x86 loses the consumer market, they will also lose the cost-sensitive server market.
x86 only owns the server market because the r&d and economies of scale paid for by the consumer market.