Intel Answers Back: Centrino Duo to Incorporate WiFi, WiMAX

Responding to last weekend's challenge from Qualcomm, which is seeking to up the ante in high-speed wireless networking with its acquisition of True MIMO producer Airgo Networks, Intel today announced that for its Centrino Duo platform, it's preparing a system-on-a-chip (SoC) that will incorporate the WiMAX technology it champions (802.16e) along with 802.11n WiFi and HSDPA.

It's called the Intel Connection WiMAX 2300 baseband chipset, and it's being described as the core connectivity provider for future notebook and laptop computers bearing the Centrino Duo logo. For anyone who asked the question after our story yesterday on Qualcomm, "Why is BetaNews making it look like WiMAX and WiFi are the same technologies addressing the same markets?" here's your answer.

Of course, it will be a tricky proposition for Intel: 802.11n WiFi doesn't exactly exist. The matter of voting on the Draft 2.0 proposal is scheduled to come up next March, but the principal proponent of the draft on the table is Airgo, which is now owned by Intel's key competitor in wireless networking.

Here's the problem: It's absolutely true that 802.11x technology addresses routers with service ranges measured in feet, and 802.16 broadband has a service area measured in miles. But in the wireless networking field, as Centrino has already proven, the "platform" is a lot wider than the computer alone.

What will matter to the company that successfully deploys the full wireless platform isn't just whether its chips are included in PCs, but whether it's also deploying the routers and/or towers providing service.

You'd think WiMAX (and, for that matter, HSDPA as well) would have the advantage. The problem is, the parties who intend to reap the most revenue from providing wireless access points to their customers and patrons include hotels, airports, and yes, coffee shops. They won't be deploying WiMAX transmitters, and they may very well wish they'd just go away. They have an interest in keeping service areas measured in feet rather than miles.

If WiMAX or HSDPA were deployed on a nationwide scale, the need for corner shops and waiting rooms to provide wireless access could cease altogether. Qualcomm produces HSDPA/WCDMA chipsets, though its strategy is to build up municipal wireless broadband as an extension of cellular networking rather than Ethernet. Still, a notebook computer producer isn't going to put both chipsets on one system. Only one chipset will fit.

Both Qualcomm and Intel have an interest in 802.11n, even if they have different interests in it - or rather, even if they have the same interest in different implementations of it. But if Intel is as successful as driving the Centrino Duo platform for medium- and long-range broadband the way it drove Centrino for short-range, computer producers might not see the need for Qualcomm's HSDPA alternative. Both players want their platform brands to be the one that appears atop the transmission towers and in the coffee shops. In short, whoever wins the platform wins all the marbles.

Make more sense now?

Intel's statement today makes its position very, very clear: With notebook computers becoming smaller, technologies that don't matter so much could get shoved aside. That is, unless they were to be incorporated in a single SoC. "As laptops become smaller, for example, they will have limited space for new technologies," Intel writes. "Integration also helps enable ubiquitous connectivity on ultra mobile PCs, consumer electronics and handheld devices that have significant size constraints for the number of cards or components."

Another key segment of today's statement is this: "The baseband chip also employs the same software for Intel's WiMAX and Wi-Fi solutions to help ensure unified management for connectivity. Over-the-air provisioning supports easy configuration and enables consumer activation of services, shifting the traditional hands on service provider business model to a direct activation one based purely on consumer purchases of mobile devices."

In other words, under the Intel system, the customer will get an opportunity to choose which services in both long- and short-range he prefers to use, perhaps through a package deal offered by participating WiMAX carriers and networks of service establishments, which could possibly include that next-generation internetworking connectivity provider of highest merit, Starbucks. Then the "platform-at-large," if you will, would ensure that, as the customer moves from place to place, the handoff between long- and short-range providers always favors the services to which the customer subscribes. When a traveler moves to a new hotel, for instance, he doesn't find himself signing up all over again for a new contract for Internet service.

This business model could lock in certain preferred brands of providers into the Centrino Duo platform-at-large, which could mean lower service fees for the end customer. It could also mean locking out the opposing platform.

Intel is saying today its design is complete - which is interesting, because 802.11n itself is not complete. But a notebook computer was demonstrated today at the 3G World Congress in Hong Kong, and an even-money bet says you'll see that same computer at CES 2007 in Las Vegas in January. Broad customer sampling of the chipset, Intel says, will begin late next year.