I just hope that, from the end user's perspective, they pull this off as smoothly as Apple managed the PowerPC to Intel transition. I'm taking it for granted that most .NET applications will work on Windows ARM out of the box, but that won't be nearly enough.
If people can't run existing native x86 Windows apps on this, it won't be thought of as "real Windows". Legacy x86-only Windows binaries will be around for a long, long time given the enormous size of the Windows ecosystem.
In two years, Microsoft will probably be competing with...
...maybe the second or third-generation iPad?
...maybe the forth or fifth-generation iPhone?
...maybe Mac OS X Lion, or later?
...whatever might follow the A4 SOC?
...and whatever else Apple is baking in the Cupertino product ovens.
Not to mention Android Saltwater Taffy, and Ubuntu Peckish Pixie, or whatever...
And then there's HP and their WebOS, RIM, and the various other vendors that are already in or are seeking to get into the ARM and phone and tablet and related markets.
And then there's the inherent delay before folks can or will adopt the new Windows platforms. How many folks that are still on WXP and thinking about W7 and WP7 (as good as those are, too) are going to leap onto WARM, after all?
Microsoft need to aim way past the existing iOS and Android and WebOS environments here and way past WP7, and they're going to be encountering a massive amount of (new) work as porting can inherently delay or derail parallel new feature development. And then the third-party vendors will need to add support.
And then will the existing customers then want to buy WARM products in quantity?
"If these ARM-based rivals can succeed in the market with their chips, they could break Intel’s near-monopoly on Windows PCs. And if Microsoft can do it, Apple probably can too."
Has me shaking my head. Apple already basically is doing that. I suppose you have to dumb tech news down in some cases, but still.
It amazes me how PR firms and money influences public opinion when I see this "And if Microsoft can do it, Apple probably can too."
Today some people believe MS invented the desktop, and they will try to convince people they started the ARM transition when all they have is prototypes when there are companies already selling ARM computers(Linux and mac)today.
Give credit where credit is due: When OneLaptopPerChild started using ARM for computers, I remember companies laughing about it: "Just a phone with a screen", "just a toy for kids", "you can't do anything with ARM", "people want power, they are never going to tolerate this". This coming from people in the industry like Intel execs.
Then Apple used it for iPad and demonstrated that people would buy it. Apple has designed and already sold a lot of this and will offer an ARM Desktop in the future. The difference is that Apple uses to show finished products, not prototypes like MS loves to do since Windows 3.11 days.
[Edit:I'm wrong here, OLPC does not use ARM jet,but AMD, but the idea was that low power comp has been seen as a toy when it could transform mobile computing]
> Today some people believe MS invented the desktop, and they will try to convince people they started the ARM transition when all they have is prototypes when there are companies already selling ARM computers(Linux and mac)today.
Today some people believe Apple invented the desktop too, indeed many elements of the Xerox Alto had been demonstrated in Doug Engelbart's mother of all demos for the NLS system in 1963, the NLS system itself being based on and inspired by the memex machine documented by Vannevar Bush in 1945. It's not always as clear as one thinks as to where credit is due. When it comes to the desktop it's as clear as mud, with PARC, GEM, Workbench, Windows and Mac OS all able to lay claim to firsts in their respective areas, which were to some extent imitated or copied in others.
Seriously though, at the risk of sounding like a fanboy, it seems clear that Apple is leading the way here.
Highly-portable touch computing is where things look like they're going, and Apple is not moving off of Intel for the Mac. They noticeably didn't pick Intel's offerings for their iOS line, which is outselling the Mac by quite a bit.
I disagree- Apple could have used Atom for the iPad. They Purchased PA Semi, and seem to be taking an interest in that space. I dont think they bought them without at least entertaining the idea that they may be able to transition this to Mac someday.
On one hand, it makes me happy to see that evolution away from the x86 ISA is possible. On the other hand, it saddens me a bit it's Microsoft driving this change - we may have better, more modern processors under the same old OS monopoly we face today.
You may need therapy if it saddens you to see Microsoft port its OS to non-Intel chips, like they used to do with NT. Who else is going drive change away from the x86 ISA?
People keep mentioning non-x86 ports of NT in the context of yesterday's news, but I think a more nuanced recollection of history may be in order.
There's one very important apparent difference between the move to support ARM and those old NT-on-RISC ports of yore: Microsoft never, for even a microsecond, took those ports seriously. They were brought into being for the sole purpose of creating uncertainty in the unix market to slow enterprise adoption of Unix on those exotic RISC architectures. NT, recall, was supposed to be "a better unix than unix" in those days. As soon as that move succeeded in neutralizing the competition, those ports became legacy.
By all appearances, on the other hand, the port to ARM appears to be intended for actual consumers — that is, not merely to freeze a market, but to enter one.
> that is, not merely to freeze a market, but to enter one
I would be more inclined to believe that if Windows 8 on ARM had, at least, an announced shipping date. As it is now, I increasingly believe Microsoft saw the threat Android on higher-performance flavors of ARM represents to Windows on x86 and decided to create some confusion by announcing Windows 8 would support it.
I have to admit they being able to actually demo it, along with Office, with credible performance surprised me. They sure came a long way since the NT on RISC of the mid-90's.
However, Windows 8 is, at least, a year away. Does it mean these products that were demoed will not ship before Windows is ready? I could run Linux on them by next week.
This is not really a fair recollection of the events. Things like NT for Alpha were not successful not because MS wasn't "serious" about it, but just by the fact that it didn't sell. Those architectures just weren't commercially viable. The products themselves were basically the same (and impervious to not being "serious") because of the investment they made in maintaining a microkernel that wasn't tied to x86.
This is closer to the mark. I remember rebuilding an NT 4.0 server on x86 after Microsoft dropped Service Pack support for Mips after SP1. We'd been sold a right lemon by the vendor who obviously must've known that about 3 months after NT4 ships Microsoft announced they were dropping MIPS. In the end, instead of buying more expensive IRIX kit or facing the political fallout from going back to Novell my then employer decided to grit teeth and go x86/NT4.
> You may need therapy if it saddens you to see Microsoft port its OS to non-Intel chips, like they used to do with NT
I would prefer some other player, like Android and the Linux ecosystem it piggybacks on, to do it. As for the therapy, I have long ago realized I just don't like the Microsoft Way Of Doing Things. The fact they have abused their monopoly to actually prevent the market from progressing away from the x86 ISA doesn't help either.
Ok. So I didn't see that coming. That is actually rather impressive.
My only "complaint" here would be that Nvidia's Tegra system, although noticably smaller than the basic Intel-setup, seemed very much to be at the same scale, i.e. big. At least not tiny by the same factor as the Snapdragon setup.
On a related note: Whoah. A snapdragon seemingly capable of running Windows smoothly.
Anyway. If Microsoft can help getting rid of the x86-legacy and move on, I'll be the first to applaud it. x86 is a butt ugly architecture and I'm surprised it has survived this long.
Its a 4 minute demo. If you've ever tried giving a business pitch for a new company to an investor, you usually have even less time. What they are doing is hitting the high points, i.e. Windows kernel on ARM, popular applications already running on the new kernel, and device integration (printer, phone, etc) with the kernel.
I'm not a big Microsoft fan, with the exception of Xbox and the new Kinect. However, you have to give credit where its due sometimes. I am blown away that the little motherboard he held at the beginning is so little. I've seen UMPC devices (my roommate had one) that ran Windows on a little piece of hardware, but very poorly. This so far looks much more impressive than that.
So while the demo looked fairly basic, you really have to step back and understand the hundreds of man hours that it took to get the core Windows kernel ported, plus drivers, plus applications. Building this stuff is not easy, no matter if we dislike the final products or not.
If people can't run existing native x86 Windows apps on this, it won't be thought of as "real Windows". Legacy x86-only Windows binaries will be around for a long, long time given the enormous size of the Windows ecosystem.