Today, Microsoft ended months of speculation and announced at CES that the next version of Windows will support SOC architectures, for x86 – and ARM.
Given the recent rise in popularity and buzz around the ARM architecture, a number of media outlets touted the potential addition of support for ARM as a necessary game-changer for Microsoft. Many of the related articles, however, forgot to mention that Microsoft has offered OSs supporting ARM for years through Windows Embedded Compact/CE and its progeny, Windows Mobile and Windows Phone.
Make no mistake – there can be a big difference in the types of systems typically running an embedded SKU versus a traditional Windows distribution. However, we have continued to observe a convergence in system requirements between the PC and embedded domains. Desktop PCs are becoming antiquated in favor of more mobile form factors such as notebooks and tablets. Meanwhile embedded devices have shed many of the traditional constraints that led to the development of the diverse solution ecosystem surrounding them. Many of these devices (a greater percentage of which are also adopting mobile form factors) now have web connectivity and more advanced system resources that now make their functionality comparable to PCs.
So if the Windows Embedded Business (WEB) group already has OSs supporting ARM, does this really change anything in the embedded market?
Short answer: Yes.
First of all, VDC estimates that the majority of WEB group revenue is currently related to the other embedded SKUs that do not currently support ARM architectures. In fact, we previously estimated that approximately 40% of WEB group revenue was related to Windows Embedded Enterprise SKUs – in other words, restricted licenses of full versions of Windows Vista, XP, or 7 etc. The group also generates a significant amount of revenue from Windows Embedded Standard, which is a componentized form of their flagship desktop OS. Given ARM’s continued push into device classes beyond mobile phones, we expect that any downstream support of ARM for other embedded SKUs will present Microsoft with another needed arrow in its quiver to maintain (or even grow) its share in the embedded market.
Similar to ARM’s attempt to enter the server market, we also expect that this move may ultimately help other embedded OSs (or even Linux) gain a greater foothold in the PC market (desktop or mobile form factor). Microsoft (along with Intel’s help) was able to maintain and reinforce its dominant position in the PC market over the years through tight control of the other system components – and thus the development of the ecosystem of supporting solutions. The embedded market certainly never had that same dynamic with a diverse set of hardware, software, and development solutions available to the engineering community. If Microsoft’s broader support for ARM spurs a greater use of ARM-IP based processors in PC machines, we expect that there will be additional opportunities in the PC domain for the other operating systems that have developed their own ecosystems and support communities for ARM architectures.
If we can take anything away from this announcement and the early goings of the tablet market (another point of focus for CES), one thing is for certain – the traditional divisions between embedded and PC systems are blurring and any solution vendor too slow to straddle these domains may lose its footing altogether.