Microsoft Demos IE10 on ARM, and It Looks Good
Blink and you'll miss it. Microsoft has demonstrated Internet Explorer 10 at the MIX11 show in Las Vegas. For around 30 seconds of the 90-minute talk, they mentioned that some of the demos were running on an ARM computer.
That's the full Internet Explorer 10, which would presumably require a reasonably complete version of Windows too, and not running on an Intel or AMD x86 processor. And running quite well too, by the look of things.
What's that strange smell coming from Intel headquarters? Oh, yes. It's fear.
NVIDIA was quick to fess up that the "demo was powered by the NVIDIA Tegra superchip." (See the video here.) They gave no specifics as to make and model, but we can guess it's a Tegra 2 Dual-Core chip, with a system-on-a-chip GeForce GPU. This is about as powerful as ARM chips get right now.
NVIDIA also hinted that it's working closely with Microsoft on the Windows 8 product.
By looking at the MIX11 videocast with a magnifying glass, we can see the chip is a humble 32-bit offering--as with all ARM processors today--and that the machine had a mere 1GB of RAM. There must be some pretty awesome optimization going on with the ARM port of (presumably) Windows 8 if it's truly able to run at speeds fast enough to give demos--especially those involving full graphical hardware acceleration, as is the impetus for many of IE10's new features.
A lot has been written about why Microsoft is porting Windows to ARM but I think this brief demo finally makes things clear: It's all about platforms, and not for the first time.
Back in the last decade I wrote extensively about desktop Linux. Often I'd speak with corporate system administrators who would tell me they'd love to get rid of Windows and put Linux on every desktop. Unfortunately, they needed Windows to run Microsoft Office.
Windows' primary worth in the corporate space was in acting as a platform. It was the horse for the Microsoft Office carriage.
With this in mind, we're all wrong to think that Microsoft porting Windows to ARM is the big deal. The big deal is making Internet Explorer cross-platform. Microsoft is showing that--just like Google with its Chrome browser and OS--it's betting that the future of computing is going to be in online services. This will involve a shift in our concept of what a computer is. And, again, Microsoft wants one of its products--Internet Explorer--to provide a compelling reason for us to keep buying Windows, regardless of which device we use.
Many players in the IT sector would like end users to consider their computer purchases in terms of platform and utility, rather than specifications. Hardware spec lists are so 1990s. The move to mobile technology is providing a means to change our thinking. For example, Apple refuses to discuss processor speeds or quantities of RAM in its iPhone and iPad documentation, preferring instead to talk about other hardware features that enhance usability. Few other tablet or phone manufacturers mention specs.
In the near future we will talk of an app running well on Internet Explorer, without making reference to the underlying hardware or even operating system. Whether IE is running on Intel or ARM will be irrelevant. We'll talk of Google's Chrome platform in a similar way. Again, it won't be relevant whether Chrome manifests itself as a browser or the entire Chrome operating system. It's all about platforms.
There's a long way to go yet, however. There's no standardized ARM platform like there is for x86 chips, and even ARM chips can vary substantially in their designs. One question asked frequently is whether Windows will run on the iPad, which after all is just another ARM-based tablet. The answer is that it probably won't, unless Microsoft hack Windows to do so. Microsoft will probably have to hack Windows for individual manufacturers of ARM-based computers over the coming years.
Just a quick side thought: The platform gold rush is starting to make Mozilla's strategy of doing little more than producing first-class browsers seem a little curious. Is Mozilla going to get squeezed out of a market that Microsoft and Google dominate? After all, the only reason Firefox became so popular was that Internet Explorer was so dire at the time, and had been stagnating for several years. Mozilla has built on this position, but has it really evolved?