Hacking Bangkok Blog

Tech and Living in Thailand's City of Angels

The Hacking Bangkok blog covers I.T. and technology in general, and my experiences working and living in the Kingdom of Thailand. Bangkok has a very long Thai name, which starts with Krung thep - City of Angels.
Bangkok sunset from my bedroom balcony

Thursday, September 1, 2011

How Microsoft could end up winning the Great Tablet Wars of 2010 - 2015


Conventional wisdom, judging from what I read in the blogosphere and even in respected techie magazines, is that Microsoft and its Windows operating system are in terminal decline, about to be eaten alive by Apple’s iPad (and iPad 2, etc).  Proponents of this view point to the exploding market for iPads versus the sluggish growth in the PC market overall, and to be fair, they do have some history on their side:  Microsoft has tried for years to sell consumers and businesses on the idea of tablet computers, even renaming them as “UMPCs” (ultra-mobile PCs) to try to undo the negative history that tablets had - due in no smart part to Microsoft’s own failures in creating a successful tablet.

So - why do I think Microsoft could end up “winning”, and grabbing the bulk of tablet sales? My view is that there are several converging trends that favor Microsoft:  notebook computers sales are still increasing rapidly, Intel has made immense progress in creating low-power x64 chips, Microsoft is releasing Windows for ARM architecture, and perhaps most importantly, Apple has shown them how to build a finger-friendly tablet OS.

Numbers first - according to figures put out by IDC and IC Insights, Apple sold about 15 million iPads in the year from its release (April 2010 - March 2011).  There were a total of 29 million sold by the end of June 2011, an incredible jump in sales numbers. Projections by IC Insights are that a total of 49 million tablet computers will be sold in 2011, with most (about 35 million based on market data I’ve seen) will be iPads. The remaining 14 million will be mostly Android tablets by OEMs like Samsung, Acer, HTC, Motorola, and Lenovo, along with a smattering of Chinese brands you’ve never heard of.  Oh - and perhaps ahalf- million TouchPads running WebOS, given HP’s recent (if puzzling) decision to make another batch of their loss-leading tablet. So clearly, Apple is selling waaay more iPads than most people - me included - thought they would.  They pretty much own the tablet market for now.

Then, there’s ye olde fashioned PCs and notebooks running Windows.  About 345 million were sold in 2010 [IDC], of which 160 million were notebook computers. For 2011, IC Insights projects that notebook sales will rise to 182 million; that’s only a 13.8% increase over 2010, but in units sold, it’s a jump of 22 million notebooks. 22 million is roughly two-thirds of the total projected sales of the iPad this year, and leaves “traditional” notebooks with sales of about 5x iPad sales this year. So my first point is that notebooks are hardly dead yet, and the endurance of their form-factor for over two decades shows me that, for many people, laptops are still their first choice for a computer.

With that said - if Microsoft, Intel and the various PC makers don’t respond to Apple (and to a lesser extent, Google and the Android tablets), well, those tablets will get more powerful, and continue eat up market-share for mobile computers. One problem Microsoft always had with making a tablet was battery-life.  When Windows XP tablets came out, and later UMPCs, the battery-life was usually pretty dismal, even by notebook standards. Three to four hours was common. The iPad gets roughly 10 hours. Apple realized that people don’t use tablets they way they use notebook computers, so they built a custom version of their OS and run it on low-power ARM chips that can go all day. 

But Microsoft, along with server-makers, has been pressuring (haranguing?) Intel to concentrate on lower-power CPUs, and on integrating more functionality onto the CPU package. Intel has responded by integrating a GPU in the same package (with the Core i3/5/7 and newer chips), putting the memory-controller and other functionality directly on the chip, and lowering the core power-consumption (in part by reducing the average transistor size down to 22 nm for the next-generation). Intel has vast resources and expertise in microprocessor design, and they’re (slowly) driving the TDP (thermal-design power) of their integrated CPU/GPU down towards the ARM architecture’s power consumption. They’re not there yet - but with every iteration, Intel gets closer, and with ever year, batteries get better by about 10% in energy density. So my second point is that Intel and AMD are making x64 systems-on-a-chip (SoCs) - which run countless applications that businesses and people around the world depend on - ever more efficient, and ever more similar to the Snapdragons and Tegras of the ARM world.

Microsoft, though, is hedging its bets for the mid-term: the next version of Windows, called Windows 8, will run on ARM processors as well as x64 silicon. To top it off, Microsoft ported its biggest non-OS cash-cow, Microsoft Office, to run on Windows on ARM. They also looked very carefully at what Apple did with the iOS user interface, and created a new UI for Windows 8. There is precedent for this. When Apple came out with the original Mac OS, and Microsoft was still pushing DOS, Microsoft quickly copied (rather poorly at first) the new graphical UI that the Mac used, and launched it as Microsoft Windows. After three versions, it caught on: Windows 3.1 for Workgroups became a “business standard”, and Apple of course sued for it. Windows 95 pushed that even further, transforming Windows into a modern (or modern-looking) OS. So Microsoft is preparing, with its OEM partners, to launch Windows 8 tablets running on the same type of chips (ARM architecture) as Apple’s iPad, but which also run Microsoft Office, and probably other heavy-hitter apps that Microsoft has convinced developers to migrate to this version of Windows. Plus, if Microsoft is smart (and they seem to be), they’ll port the .Net framework over also, meaning that a ton of existing business apps will run on these ARM tablets also. That’s a compelling advantage for a lot of companies who buy tablets for their employees. It’s even pretty compelling for me, a software developer.

Finally, I could comment on Microsoft’s total failure to create a finger-friendly tablet OS, but anybody reading this probably already knows that sordid history all too well .  But - as I mentioned - Apple essentially created this market, and has shown how to create a user-friendly experience on a tablet. Just as Google/Android leveraged that knowledge (if “leveraged” is the right word) to vault Android to the market-leading smartphone platform, Microsoft is going to try to leverage Apple’s experience with the iPad, plus their own experience, to create a tablet-friendly Windows 8. From the early looks people have had, they’ve succeeded; Windows 8 looks good, and works well.

As the SoCs that ARM designs, and that Intel and AMD design and build, get more powerful, tablets are going to be able to do more and more serious work. It reminds me a bit of how notebook computers themselves were originally unsuitable for some jobs that you needed a “real” PC (that is, a desktop PC) to do, including gaming.  Modern notebooks can do pretty much everything a desktop can do, including play video-games (yes, I know, mobile GPUs aren’t nearly as powerful as the latest Radeon or Nvidia cards, but they’re good enough for most people). And tablets might (probably will?) follow the same route. The important thing is, this benefits Microsoft, since Windows is a large, full-featured OS compared to iOS, and requires a bit more oomph under the hood. 

What happens in five years, if ARM-based tablets have more computing power than high-end desktops have now?  Will Apple migrate Mac OSX over to ARM as well? Or merge Mac OSX with iOS (itself based on Mac OSX)? Microsoft will already be there, with Windows on ARM, and best-selling application suites running on it. What if Intel does manage to push x64 power-consumption down to tablet-friendly levels? Microsoft is there also, with Windows 8 for x64 running not only Microsoft Office, but almost every app ever written for Windows.

In short - if Microsoft doesn’t screw up (and granted that’s an “if”), I think they could end up doing to Apple what they did two decades ago: mimic their UI/UX in Windows, and let the giant OEMs put out hardware that competes with Apple spec-for-spec. Apple knows this, and isn’t going to just sit and roll over - they’re trying to tie every cloud-based service they can think of to iOS. But the big companies that drive a lot of PC sales aren’t generally early adopters, and by the time they’re ready to buy tablets for most of their mobile staff - in 2013 or 2014 - tablets that are every bit as capable as iPads, but which run the main applications they use and have developer environments (like Visual Studio) that their in-house developers know, will be available from the vendors they have long-standing relationships with. That’s going to be a very compelling argument to go with Windows 8 (or 9) tablets.
-        Kirk Davis

4 comments:

THompson said...

Kirk,

The true challenge for Microsoft is embodied in this observation by you:

"Microsoft is preparing, with its OEM partners, to launch Windows 8 tablets running on the same type of chips (ARM architecture) as Apple’s iPad, but which also run Microsoft Office, and probably other heavy-hitter apps that Microsoft has convinced developers to migrate to this version of Windows."

Screen real estate has significantly decreased to the point where it is not practical to have more than one window shown at a time, and there simply isn't room for very many widgets, tools, and menus. Microsoft is going to have to figure out what to leave out of the OS and each Office program and how to best arrange the remainder. I have very little confidence they can successfully do this. They have never been good at subtraction, in spite of the fact that it has been obvious for the last few iterations that the Office suite could use quite a bit of it.

Thompson

Anonymous said...

Kirk,

Very interesting read, and well written considering the source. ;)

I've been trying to get Stacy onto an android phone but she has been resistant of a smartphone at all, but I had her playing games on my phone which she quickly got addicted too. Then she was playing around with Ryan's IPAD2 and oohing and ahhing. A good tablet that can run excel and all the fun games, with web browsing is what she wants. (She spreadsheets everything)
I hope Microsoft can pull it off, just to give a real competition to Evil Apple.

Sean

Kirk Davis said...

@THompson - screen real-estate (in terms of pixels, at least) on most of the tablets that are out now (like the iPad), Xoom, and Galaxy Tab 10.1, matches or exceeds that on most laptops five years ago, which were generally 1024x768. Some of the newly-announced Android tablets are 1280x800, which is better than a fair number of notebooks (used for Excel) today.

So while I agree that running the current Office interface on a tablet with below-par resolution (like some of the cheapo Android tablets using 800x480), using it on a 10.1" tablet at 1280x800 or the higher-resolutions that I think are inevitable (1920x1080 is absolutely coming to tablets in a couple years) will be fine.

You raise a good point about the entire Office "ribbon" interface, though; it's not very finger-friendly, although my guess is that keyboard/pointer accessories are going to be the norm for creating new spreadsheets and documents (since typing is much faster on a physical keyboard for most people), while users would just grab the tablet itself - sans accessories - to "consume" content: watch movies, show presentations to customers, browse the web, etc. The onscreen keyboard and voice-recognition (as both Android and WP7 have now) will suffice for composing emails and short memos, etc.

That's just my take. I expect quad-core Windows 8 tablets with 4 GB of RAM, 128 GB SSDs, and high-resolution screens to be available by mid-2013 at the latest.

Anonymous said...

Interesting analysis. I think MS has fallen too far behind to win, but it would be good to see them provide Apple and Google with some competition. Their best shot will be for enterprise tablets, where Office and Windows will be advantages and there will still be some decent margin in the units.