By now, just about everyone knows that Google Android isn't just an operating system for mobile phones. T-Mobile was outed today by the New York Times as planning a number of Android-based devices for the home. Plans are going forward for Google's Linux OS to be put on ARM-based Mobile Internet Devices(MIDs) and, more importantly, netbooks in the next few months.
ARM based? (Try to stick with me here.)
If that was your reaction, you aren't alone. If you've even heard of ARM, you know that they design embedded chips and chips for all kinds of small mobile devices, phones and iPods ... but until now, not laptops. In fact, you probably didn't know that ARM based chips far outsell Intel chips worldwide.
Why would Google's hardware partners and perhaps even Apple choose to make a laptop with an ARM based chip?
ARM's new Cortex A8 based processors have roughly the same horsepower per cycle as Intel's Atom 270-280 chips but run at a fraction of the power and cost a fraction of the price. Their next generation chip designs, the A9 series, are even faster and more miserly on power. This is exactly what netbook makers want.
They can run Web browsers and Adobe Flash applications with ease and also do very well with OpenGL based graphics. They can run Linux, MacOSX, Symbian and even WindowsCE/Mobile fairly easily (but they don't do regular Windows).
ARM doesn't produce their own chips. They license the designs out to other hardware vendors like TI, Samsung, Qualcomm, NVIDIA, maaaybe Apple and even Intel. The hardware manufacturers are responsible for marketing them to hardware vendors and getting them into devices. The problem here is that all of these different hardware vendors are going to compete with mighty Intel for the hearts and minds of consumers, not hardware engineers. They are out of their league.
These ARM Cortex A8 processors are already appearing in high end phones and MIDs like the Palm Pre, Open Pandora and Archos' new phone. In these devices, the processor is called the TI OMAP 3400 series.
So if you aren't an ARM processor geek, you are probably starting to get a little lost here. This is the problem for ARM.
The average consumer is going to be seeing Netbook A based on the Intel Atom chip. Netbook B is going to be running something like the NVIDIA Tegra based processor. There are 500 different netbooks on the market right now; 95% of them use the Intel Atom chip. Most consumers, given that choice, would pick Intel because they know the name and the Atom brand.
That's why ARM really needs to start thinking about consolidating their hardware brands and come up with an umbrella idea that its partners can market. The idea needs to address the fact that ARM is in more machines than Intel, uses less power and costs less money but is still a superior product.
Think brand isn't important in processor sales? A few years ago, when Intel was playing with Itanium, AMD slipped in a superior 64bit processor line that flat outran Intel for a few years. Intel's name and brand kept them afloat until their engineers caught up with and eventually surpassed AMD's designs. Without Intel's superior name and "Intel Inside" program, they might have lost out to AMD. Now AMD is close to bankruptcy while Intel is flying high.
ARM needs to learn a lesson from this. They have a solid name in "Cortex". It is the core of the machine which is the brain like the cerebral cortex. It could be the next "Pentium"... or "Cortex Inside". But they need to really work on this and spend some money delivering the message to consumers. They are about to go head to head with mighty Intel in the battle for the netbook/MID.
They can't afford to go in unARMed. (I'll be here all week folks!)
This story, "ARM-Wrestling Competition for Chip Makers" was originally published by Computerworld.