Via Weblogsinc comes news that TeleCIS is working on a chip that will combine support for Wi-Fi and Wi-Max.

The chip is due in the second half of 2005. A lot of people are very excited about it.

I'm not one of them.

Here's why:

The assumption here is that Wi-Fi and Wi-Max are somehow competitive, that Wi-Max will be a carrier technology only in 2005, and will then filter down to the LAN level.

TeleCIS plans to mix the two, making them directly competitive. This is the kind of thinking that you get when no one is applying any vision to a space.

And I'm talking here about Intel. Intel needs a hit in the communications space, having missed the original Wi-Fi boat because it failed to do business in the way manufacturers demanded. Its chips were good, its business process just stank to high heaven, and they haven't gotten the smell off their clothes yet.

See, we lost Wi-Fi, but we'll win Wi-Max. Thus you need to overlay because the two will compete.

Nonsense.

Wi-Max is designed to send big hunks of data over vast distances. It's a wholesale technology. Wi-Fi is designed to send somewhat-smaller chunks over much smaller distances. It's a retail technology.

Wi-Fi continues to move ahead. The next version, 802.11n, will run data at over 100 Mbps. There is going to be a crying need for backhaul, not just long-range backhaul but medium-range as well.

If you can't get your collection of Wi-Fi traffic out to a competitive fiber node you're going to be at the mercy of the Bells, who sell T-1s through an eye-dropper, with an aim toward putting you out of business and taking it all over for themselves to subsidize their wires.

That's where Wi-Max comes in. Wi-Max can do that job, and do it quite easily. The trouble (for Intel) is that, when it's doing that job, the volumes of chips needed is relatively small -- you don't need as many wholesalers as you need retailers.

There are some things Intel could do to move things along. They could encourage the re-emergence of Wireless ISPs, who will use both Wi-Fi and Wi-Max. (But that business is held by Alvarion.)

They could also make a new play for the Wi-Fi market based on a having a robust scalable platform, on cognitive radios, on mesh networking, and on Always-On applications. That's what I would do, frankly.

Someone is bound to do this. It's a shame if Intel won't. But, as we've already seen, the marketplace's response to Intel is a bit like that of Eliza Doolitte's to Henry Higgins. It can do very well without you.