SAN JOSE, Calif. – Mark January 5, 2011 as the date of the big PC earthquake. That's when Steve Ballmer said the next version of Windows will run on ARM and Jen-Hsun Huang made Nvidia the first chip maker to say it will deliver a soup-to-nuts family of ARM chips for computing.
Old structures have been shaken to the foundations. What will emerge when the smoke clears is still anyone's guess. One thing is certain, all forms of computing from notebooks to supercomputers will see lower power, lower cost versions based on integrated ARM SoCs.
That fact alone has long term repercussions for everything from the emerging tablet market to billion dollar data centers starved for electricity. It will even bring home PCs to modest dwellings that would never otherwise have seen them.
These issues have socio-economic implications that will ripple far beyond the history of an industry. But the implications for the electronics industry are also significant.
Ultimately Intel, the world's largest semiconductor maker, will not be able to sustain average selling prices that are dozens and sometimes hundreds of dollars over those of other chip makers. It will find itself unable to sustain the vertically integrated business model in which it is the world's largest operator of chip fabrication plants.
There is some inherent danger for the whole semiconductor industry here. What happens when no one company has the motivation and financial clout to plow through the enormously difficult tasks of delivering new chip processing technologies?
Intel helped pioneer extreme ultraviolet lithography, the long delayed printing technology still expected to drive the industry to ultra fine chip patterns. It is also driving the shift to larger wafers.
The need for many small players to collaborate on the future of chip technology will complicate the financial and social aspects of technical tasks that are already nearly insurmountable. Expect slower progress right at the time when CMOS technology hits atom-sized limits in physics and demands radically new approaches.
And what about systems companies? Hewlett-Packard is the world's largest user of semiconductors, most of them going into x86 PCs and servers often defined at the motherboard level by Intel and AMD.
PC makers such as HP, Dell, Lenovo that have slashed their ranks of chip and board engineers will have to re-grow those capabilities again. They will need to evaluate perhaps a dozen competing SoCs and define board designs based on them to differentiate themselves. They may even need to design ASICs again—mon dieu!
Taiwan's emerging OEMs will have an advantage. They have already nimbly crossed over to designing and making ARM-based handsets as well as x86 PCs. The Windows-on-ARM shift is their chance to use their engineering prowess to break out of the role of silent partner to big brand companies such as Dell and HP.
All this change won't happen overnight. The PC earthquake of Jan. 5, 2011 is one that will play out in slooooow motion.
It could take Microsoft years to deliver a solid version of Windows on ARM. Windows itself was largely a flop until it hit version 3.0, and even once it became an industry behemoth, Microsoft was still apt to fall flat on extending its franchise. Remember Vista?
That said, Microsoft had to do this to stay relevant. It risked losing its core franchise to Linux variants such as Android and Chrome OS already climbing up from smartphones and into notebooks and even into ARM-based servers which are getting a thumbs up from the likes of Dell and IBM.
Once Microsoft delivers, the ARM chip makers need to step up, too. They are used to designing 8- to 32-bit embedded SoCs for handsets and a plethora of other systems. None have done a 64-bit design--ARM doesn't even have a core to base such a design on yet.
What's more, none of the ARM SoC makers have had to deal with supporting the complexity of a software stack like Windows with its millions of lines of code. Just as daunting is the need to support the world of PC peripherals with their often poorly written drivers, a thorn in the side of Microsoft for years. Nor have they had to deal with supporting chips that might sell in volumes of tens of millions, a problem they would probably welcome.
In short, there are implementation and execution issues here that will take a decade to play out. But the shift officially has begun. Get ready.