Other than the death of the Start button and the Metro UI, the single biggest change in Windows 8 is that it’s now a fully paid up member of the touch-first ARM ecosystem. After 20 years of being almost exclusively x86, this tectonic shift was triggered by two key factors: ARM is cheaper than x86, and it’s also more power efficient. With smartphones and tablets enjoying meteoric sales, and the x86 PC market stagnating, Microsoft doesn’t really have a choice in the matter: Embrace cheap, mobile computing, or perish.
But then along came Medfield, Intel’s first x86 SoC, with performance and power characteristics that are comparable to the Cortex-A9 parts found in almost every smartphone and tablet on the market. It has been said numerous times that Chipzilla, with its monstrous profit margin of around 25%, could never price its parts to compete with ARM manufacturers — but judging by the price of the Xolo X900 ($420), the first Medfield-powered Atom phone, it seems Intel is finally ready to play ball.
Now, it isn’t 100% confirmed yet, but it’s very likely that by the time autumn rolls around, there will be Cover Trail-powered Windows 8 tablets. Clover Trail is the dual-core, now-with-a-souped-up-GPU successor to Medfield. We don’t have any benchmarks to share, but we wouldn’t be surprised if its performance matches or improves upon Qualcomm’s Krait, the current leader of the ARM pack.
In other words, come the release of Windows 8, there will be a range of ARM and x86 tablets that are the same price, the same weight, the same performance — but not the same operating system.
Therein lies the crux: Windows for ARM (or Windows RT in marketingese) is completely crippled compared to Windows 8 x86/64. Windows for ARM does not support Storage Spaces, Windows Media Player, BitLocker, or file system encryption. On the enterprise side of things, Windows for ARM does not support domains or group policies — your WoA tablet won’t play nicely with your office network at all. Windows for ARM won’t act as a Remote Desktop host, either.
Most importantly, though, Windows 8 ARM tablets won’t be able to run x86/64 software — while, on the flip side, every legacy program will run on an Atom-powered tablet, as will every new Metro app. At launch, while there might be a few thousand apps for Windows 8 ARM tablets, there will be a few million apps for Atom tablets. It’s hard to overstate how important this difference is. For the same price, you can either have an x86 tablet that replaces your desktop, netbook, laptop, and iPad, effectively ushering in age of ubiquitous mobile computing — or you can get a Windows for ARM tablet that replaces nothing, and will probably have significantly less apps than the iPad.
It makes me wonder why Windows on ARM was developed in the first place. I mean, giving consumers more choice is usually a good thing — and there will certainly be some bargain basement ARM tablets that are cheaper than their x86 counterparts — but that’s a steep price to pay for fragmenting the world’s most popular software ecosystem.
It will be very interesting to see how the OEMs handle the x86/ARM schism. Samsung, for example, will almost certainly sell a Windows 8 ARM tablet and a Clover Trail tablet and Core i3/i5/i7 Ivy Bridge tablets. Somehow, OEMs like Samsung will have to differentiate between ARM and x86 tablets that have almost identical feature sets, and yet one is completely new and untested with relatively few apps, and the other is tried and tested and has the support of the largest app and developer ecosystem in the world. I don’t envy Windows RT.