4 Comments

Summary:

WiMAX was a big flop for Intel, but instead of giving its wounds a quick lick and chasing after the easy LTE money, Intel is investing in yet another 4G variant, time division-LTE. Why? Two reasons: Qualcomm and China.

Intel is back in the mobile broadband saddle, but it’s trading its rusty WiMAX spurs for shiny new LTE ones. But when I say LTE, I’m not referring to the same long-term evolution technology that Verizon, AT&T and most of the world’s operators are using today. Instead Intel has taken a peculiar interest in time-division-LTE, a variant used by Clearwire and most notably carriers in China.

Intel and Huawei have agreed to set up a joint lab in China to test their respective TD-LTE infrastructure and device silicon. These kind of interoperability testing agreements are a dime a dozen, but this one is noteworthy. It shows that Intel is not only keen on carving a new mobile broadband business for itself, but it has a specific geographic market in mind: China.

Back to the drawing board

When 4G standards wars broke out, Intel placed all of its bets on WiMAX in the hopes of breaking what it considered a telecom stranglehold on wireless technology and replacing it with a more Silicon Valley – and Intel – friendly networking standard. If you haven’t figured it out already, it bet on the wrong the tech.

But now that the dust has settled in the 4G wars, it’s become readily apparent that the clear winner isn’t just LTE, but specifically the frequency-division duplexing (FDD) version of LTE that most of the world’s operators have embraced. In fact, TD-LTE is being pursued largely by the same operators that championed WiMAX (It’s a function of the unpaired spectrum those carriers own, not some sense of misplaced loyalty).

After falling flat with WiMAX, why is Intel making the same ‘mistake’ by championing what’s basically WiMAX Jr.? Why not follow the big money and make FD-LTE silicon and equipment just like everyone else?

My guess is that Intel realizes that the FD-LTE ecosystem is going to look a lot like the ecosystem that drove 3G, and that means it will be dominated by wireless chipset maker Qualcomm. Intel will have a role to play in FD-LTE, however: It bought Infineon specifically to become a player in the wireless modem arena – with one eye cast on LTE in the future. Infineon propelled Intel into the No. 2 slot in the wireless baseband rankings, but it’s a distant second. Qualcomm makes nearly half of all radio chips embedded in mobile phones, tablets and other devices today. And in the rapidly field of mobile applications processors, Intel’s Atom chips made only the tiniest dent.

But in TD-LTE Intel can make a much bigger impact. TD-LTE and WiMAX aren’t that different technologies and the enormous investments Intel made in developing its WiMAX product and patent portfolios will come to bear (Intel bulked up its LTE intellectual property by buying Aware’s LTE and wireless LAN patents for $75 million last week). The relationships it built with carriers when it led the WiMAX Forum will also be a factor since those same carriers are now the ones deploying TD-LTE.

Intel brushes up on its Mandarin

Then there’s China. When it comes to wireless, China likes to be different. It comes up with its own wireless standards, it plays by its own spectrum rules and it questions what the rest of the wireless industry accepts as gospel. As the world’s single largest mobile market, China can afford to be contrarian, and Intel for some time has been tapping into that country’s proclivity for doing things different.

It’s no coincidence that the first smartphone Intel announced with an Atom processor is destined for a Chinese operator. Intel also has wireless infrastructure ambitions, aiming to overturn the established base station architecture by powering the wireless network with generic computing platforms called Cloud-RANs. So far the only operator to work closely with Intel on Cloud-RAN is China Mobile.

China Mobile also happens to be building the largest TD-LTE network in the world, and it has the single largest number of subscribers in the world. Maravedis-Rethink projects that by 2016 there will be 89 million TD-LTE connections in the world compared to 350 million FD-LTE connections. While those FD connections will be spread across the world, most of those TD connections will be in China. If Intel can embed itself with Chinese operators it might finally become a mobile broadband force.

Intel image courtesy of (CC-BY_SA) Flickr user huangjiahui. WiMAX image courtesy of ImagineWiMAX. Shanghai image courtesy of Flickr user Robert S. Donovan

  1. Raymond Moncada Tuesday, May 1, 2012

    Hi, very good read. It is interesting to see how intel working to the mobile market since they are loosing in the table world to nVidia.
    Question: “Maravedis-Rethink projects that by 2016 there will be 89 million TD-LTE connections in the world compared to 350 million FD-LTE connections.” Where did you get this fact?

    Share
    1. Kevin Fitchard Tuesday, May 1, 2012

      Thanks Raymond,

      Maravedis-Rethink put out a press release last week with some data points. The have some info on their site as well, though they seem to have revised their numbers since this came out: http://www.rethinkresearch.biz/free_page_view.asp?crypt=%B3%9C%C2%97%8E%82%86%AF%BC%C2%88%97ksw%88%8B

      Share
  2. Rupert Baines Wednesday, May 2, 2012

    This rather dramatically over-states the difference between TD and FD.

    First, WiMAX did support FDD. Or rather, 802.16d actually did (and was supported by Intel’s chips); 16e it was on the ‘ToDo’ list.

    For LTE something like 95% of the system is identical for FDD and TDD and every manufacturer will support both in their chipsets. eg look at QCOM’s MDM9225 which supports LTE FDD and TDD (as well as TD-SCDMA – perhaps more relevant if you are focussing on China-specific technology).

    The surprise about TDD is given how widespread it is (Clearwire has more LTE spectrum than ATT or Verizon; in Europe the 2.6GHz band will be shared almost equally FDD & TDD) why people think of it as some China-specific niche.

    Share
  3. dick kessler Wednesday, May 9, 2012

    Intel seems to go through these bipolar events in their history: they dominate the pc cpu market and every few or more years someone there thinks that the good times will someday end-a pretty fair statement-though someday takes years. so someone gets a budget casts about for something that looks like a diversification-wimax certainly is/was one of those; medical systems (no one seems to know what this means-sensors on old farts like me to detect when I have fallen or something else yet to be defined), and now of course td-lte; I suppose if you believe the china mobile subscriber numbers and if you are as paranoid about survival as famous intel leaders have been then you take notice and make it your favorite tech of the year/month/decade, I suppose this all is reasonable except for one terribly difficult thing-intel has never made a success of any of these favorites to equal cpus or significantly lessen their dependence on cpu revenue. so what to make of it-nothing-in fact if the intel technology seal of approval (cue the tv music here) is a leading indicator RUN RUN RUN for the tech exit
    signed dick- old enough to remember the very first intel technology love in-bubble memory

    Share

Comments have been disabled for this post