When asked “why are you supporting WiMax?” following his appearance at the San Jose Intel Developer Forum (IDF) earlier this year, Intel Corp.s Senior Vice President and Chief Technology Officer Pat Gelsinger responded “Money”.
The question isnt “does Intel want to make money?” Of course it does. What Europeans are asking is “why does Intel feel it can make money from WiMax?”
The decision by Nokia Corp. to drop out of the WiMax forum was followed by the decision to rejoin. A lot of fluff and flannel has been offered to explain Nokias change of heart. But at the back of it is a simple question, one that is bothering all European (and far Eastern) telecommunications players: Just what is Intel trying to achieve?
First off, nobody really expects Intel to make, or supply, or install WiMax infrastructure equipment. Intel wont make masts, transmitters, fibre backbone links. It wont own, or operate construction companies or installation contractors. What Intel sells is what Centaur Technologys founder Glenn Henry once called The Black Hole: the big processor that sucks everything into itself.
Two things stand out in this. First, the success of Centrino. Intel is selling its Pentium M (and now the new version) processors, and the supporting chipset, for substantially higher margins than its previous Pentium III notebooks could command, and its all Wi-Fi hype that allows this. Wireless is magic. Intel wants more wireless magic.
But wireless doesnt really work without broadband. For wireless to be essential in a notebook, wireless has to be ubiquitous. What Intel wants is a bigger Total Available Market for wireless. To achieve that, it needs a bigger broadband market. And broadband is stalled. The worlds big telcos are dragging their feet about replacing copper with fibre and are simply refusing to take copper or fibre out to the rural areas unless they receive some kind of political subsidy.
On paper, WiMax looks like a great way of doing this. It will reach rural areas with pretty high-speed wireless backhaul. And once there, it will also (in a couple of years) be able to transmit directly to individual notebooks.
Thats on paper. In reality, thats nonsense.
The range of WiMax in a built-up area is going to be only slightly greater than Wi-Fi unless you pump the transmit power way up. You may get away with that in North America, but it simply isnt going to be allowed in Europe and much of the rest of the world. There are both good and bad reasons for that. And, of course, people will argue about the accuracy of the statement. At a recent Merril Lynch seminar, however, most engineers pretty much supported the view of people like Vodafones Alan Harper (director of strategy) and others, including those at SR Telecom: You cant break the laws of physics.
WiMax, under Intels plan, replaces fibre and also replaces hot spots.
How can Intel make money out of that? Only by persuading the whole world to go to WiMax and roll out universal broadband or (just as good) by persuading the worlds foot-dragging telcos that if they dont start putting fibre in the ground, other technologies will replace them.
And if that is the plan, the return of Nokia, plus the recruitment of British Telecom, shows that it is working.
But I think theres more to it than that. Next week, Ill look into the issues of frequency allocation and the restrictions on channel width and Ill be recalling Intels foray into the modem business 15 years ago. I think we should look into Intels “soft radio” plans for clues to where Black Hole economics will strike next.