Intel's platform upgrades are well-planned affairs. With one click of the calendar, Santa Rosa is now Centrino Pro, more than 240 new models of laptop appear on the market and another chunk of the endless roadmap moves to the rear-view mirror. Already, we're looking forward to Montevina, the code name for the next round: 45nm processor, better graphics, lower power consumption and better wireless.
Yet in Europe, we're only going to get three of those four. Shiloh, the wireless component of Montevina, will add WiMax to Wi-Fi — but in Europe, as Intel has already admitted, WiMax infrastructure is going to be a long time coming. Instead, the chances are good that 3G with its HSDPA and HSUPA extensions will be firmly established long before WiMax turns up in force. It's also entirely possible that the economics of competing with an entrenched 3G system will put off WiMax operators altogether. The right thing to do would be to have a 3G option and, if Intel doesn't fancy making the radio itself, to co-operate with an expert. Intel takes care of the power and integration issues, the things that make Centrino special, and the partner does the wireless and conformity testing.
That's such a good idea that Intel announced exactly that. Dadi Perlmutter, head of Intel's mobile group, told the world at September 2006's IDF that the company was working with Nokia to put 3G on Santa Rosa. Now, six months later, the platform is out and 3G is nowhere to be seen, not on circuit board nor roadmap. "Not sufficient return on investment," says the company.
There are three ways to read this: that there are already enough 3G options out there to dissuade Intel and Nokia from investing; that the two parties got to bickering about who'd pay for what; that the last thing Intel wants is to give 3G a leg up in the markets where WiMax still has a chance. The first option doesn't bear comparison with the Centrino Wi-Fi strategy: there are lots of alternative Wi-Fi chips out there, but Intel sells on integration, power management and certification. How the Intel-Nokia deal was financed, and what happened thereafter, will probably remain a mystery: neither company is bad at keeping secrets, especially about unlaunched failures. As for the last option: Intel has a lot more invested in WiMax than 3G — motive enough to spy a cuckoo in the chipset.
Whatever the real reason, it's leaving Europeans out in the cold — and leaving a gap in the roadmap that a wily competitor could and should invade. Next year's launch may be bumpier than Intel would like.