Google's Android - software development easier because of the underlying instruction set and processor optimizations

Intel is running scared, as PC sales continue to slow and successful mobile platforms shun Intel chips. Intel's response has been to buy companies such as Wind River and McAfee in unrelated industry segments, then promote the Intel "ecosystem" as a consistent platform that will make software development easier because of the underlying instruction set and processor optimizations. Or so Intel software chief Doug Fisher argued unconvincingly today in a presentation to journalists. Reality begs to differ: Apple succeeded quite well in adapting the Intel-based core Mac OS X to the wholly different chip architecture of the ARM processor used in apple iPhones and apple iPads. You could argue Apple is exceptional, as it had ported Mac OS X previously from the IBM PowerPC architecture. But it's not just Apple that's willing to leave the familiar but pricey, power-gulping Intel architecture in the new era of mobile devices. Google's Android runs on ARM processors, as do RIM's BlackBerrys, Nokia's various (mainly Symbian) devices, and devices running Microsoft's Windows mobile Phone. Even Windows 8 is supposed to run on ARM chips for tablets, though at the price of not running legacy Windows (7 and earlier) applications. Intel has ported Android to Intel chips, and a really bad port is available from View Sonic proving that the chip architecture is not the gating factor. After all, Google Android apps run on both Intel- and ARM-based chips, and so far there's no demonstrated advantage of Android on an Intel based tablet rather than an ARM-based one. So what exactly is the Intel architecture or ecosystem advantage that Fisher says lies at the heart of Intel's software strategy?

The portfolio of software vendors that Intel has assembled are credible providers, but they really offer no synergy with each other or with Intel. Some, such as McAfee, are focused on yesterday's problems, notably the security shortcomings of the Wintel platform. That's not Intel's or Microsoft's fault -- that platform was created in a very different world. Mac OS X, apple iOS, and Android have their own security issues, but their architectures are inherently more secure (well, maybe not Android). They leverage knowledge and techniques that debuted in Unix, a world not starting on Intel architecture and had from the get-go more concern over mission criticality. If McAfee and its competitors become critical products in mobile phone devices, that would simply indicate a failure of the mobile platforms (client and cloud) themselves. Another sign of a "not getting it" strategy was clear in Fisher's evasive nonanswer to the future of the open source MeeGo mobile phone operating system that Intel (and before it, Nokia) mismanaged. It makes perfect sense for Intel to want to expand outside of the chip business. But it's not clear that it yet knows how, at least when it comes to software.