Wintel is dead, long live ARMdroid

Posted by Charbax – January 21, 2011

CES 2011 is the time when all the major Wintel PC vendors announced major ARM Powered projects to redefine the core of their business. Almost all of Microsoft's PC vendor partners abandoned Microsoft exclusivity; and Microsoft's next-generation operating system has abandoned its exclusivity with Intel.

1. HP who makes 19.8% of the Wintel market (*1) is doing their attempt at controlling an ARM Powered platform in trying to make the Palm WebOS (I think they will also have a plan B using Android and eventually go with that). HP just said WebOS is to be designed for Tablets and Laptops also.

2. Acer who makes 18.5% of the Wintel market showed the Acer Iconia Tab A500 (possibly the slickest ARM Cortex-A9 Tegra2 tablet yet), they also just showed one of the top designs for a 7" Android Tablet using Qualcomm's 1.2Ghz Dual-core MSM8660, Acer is already selling many among the best value Android smartphones too.

3. Dell who makes 11.5% of the Wintel market has the Streak and is ramping up with a 7" Looking Glass and has plenty smartphones announced.

4. Lenovo who makes 8.7% of the Wintel market tried to announce ARM Powered Laptops and Tablets last year, but may have been intimidated or otherwise delayed (probably waiting for software), but they are back with a nice looking Android UI layer for their Qualcomm Powered U1 Tablet. Lenovo has also just announced that they have created a separate division to focus on smartphones and tablets.

5. Toshiba who makes 5.3% of the Wintel market is redoubling efforts in ARM Powered tablet and is already the first to have put an awesomely nice looking ARM Cortex-A9 Laptop hardware design on the market in the AC-100 (currently only in few European markets and not fully backed up with the productive software yet).

6. Asus who makes 5.2% of the Wintel market has shown an impressive 10" Tegra2 Honeycomb Laptop/Tablet hybrid and is also the first to show a product with Qualcomm's next big Dual-core Snapdragon processor. Even though Asus already had the best looking Snapdragon Android Laptop at Computex 2009 which mysteriously disappeared minutes after being mistakenly shown at a partners event.

7. Apple who makes about 3% of the x86 based Laptops in the world using their Mac OS, has based their whole wealth, generating most of their revenues and profits from their ARM Powered devices, especially the iPhone. Thanks to ARM, Apple has become the second largest company in the world after Exxon Mobile.

What made Microsoft decide to show Windows for ARM now?

Microsoft and Intel used to be the best buddies of Silicon Valley and rely on each others bloat algorithms to generate a steady flow of consumer demand, keeping PC prices high and keeping outrageous profit margins all along the way.

At one point, Intel somehow decided to make Moblin just in case OLPC went out of control against their Netbook. That merged into Meego with Nokia who also was desperate to get a major partner in platform.

That got Microsoft mad (*2). Steve Ballmer probably broke some windows in Redmond throwing chairs when Meego got announced. And even though Microsoft has been experimenting with full Windows on ARM for years, they decided to bet a pretty large farm on optimizing Windows for ARM all the way, to the point they can announce deep kernel integration and hardware acceleration partnerships with the major ARM Processor providers. Partnerships that could go from current Cortex to future designs and planning beyond.

My theory is, Microsoft could release full Windows 7 on ARM today for Cortex-A8 if they wanted (they probably even had XP running on ARM in their secret R&D labs a few years ago). Microsoft showed the full Microsoft Office fully working full speed on ARM already at the CES keynote. The rest of the whole x86 app and driver porting business can be fixed with cash injection (nothing that a few billions can't accelerate), tell me in the comments if I am wrong, but that whole porting of apps to Windows on ARM could possibly be done with a bunch of free re-compiling automation tools that Microsoft could release and a new Windows on ARM Application Marketplace.

For now Microsoft is going to take their time, they had to announce Windows on ARM now to allow the industry to prepare and design for it. But they won't actually ship the software until they really have to. And they might as well plan ahead to have it ready when all the super fast ARM Cortex-A9 processors with Laptop/Desktop optimized fast I/O memory bandwidth architechtures are on the market. That time could be approaching fast.

Will Intel make an ARM Processor?

Right after Steve Ballmer's keynote, while people were walking out of the Hilton keynote hall, I asked Shmuel Eden (Intel Vice president, GM PC Clients group) if Intel is going to licence an ARM processor now. He stood up for a second, I don't think he thought I was joking and said "Why should we? We actually think we have a good low power system".

For that second he stood there, my theory is he instinctively was wondering if I "knew something" but I had to tell him I was just joking (even though I was not), then he said "well if you are joking that's okay".

I think Intel can afford to put a few thousand engineers on making the best possible ARM processor they can, based on Cortex-A15 or their own custom ARM compatible designs. Intel can afford to make both ARM and x86 at the same time. Just put the choice out there and let the OEMs and customers decide which type they want to use.

This increased competition in the processor market is the biggest threat to Intel's very large profit margins, which is probably why they aren't interested in encouraging it to develop even faster.

*1 Those numbers are on wikipedia's list for the 4th Quarter 2009 Market share of leading PC vendors, let me know in the comments if you know of more recent numbers.

*2 Same thing happened between Adobe and Microsoft. When Microsoft started to make Silverlight, that made Adobe mad because they were purposefully not hardware accelerating Flash on other platforms to keep Wintel empire steady, so that triggered Adobe shift focus to optimizing Flash on ARM and Android platforms.

More reading:
http://blogs.hbr.org/cs/2011/01/the_fall_of_wintel_and_the_ris.html
http://www.asymco.com/2011/01/06/this-is-the-most-exciting-ces-ever/