While drinking coffee this morning, we stumbled into a Wi-Fi Weblog article dubbed "WiMax Gaining On DSL and Cable Modems", which proudly proclaimed WiMax was "gaining in popularity and is poised to become the third most used high-speed internet access technology after DSL and cable modems". That's damn impressive for a technology that hasn't been released yet - and may not make a serious impact on US broadband until 2009 - if ever.

This latest wave of optimism comes from a Yankee Group study that more reservedly predicts WiMax could compete with DSL and cable many years from now. "The largest impact on WiMAX market acceptance will be when silicon is embedded in portable devices, such as laptops, removing any upfront investment in CPE (Consumer Premises Equipment). This will likely happen in 2007 when 802.16e silicon is produced," says one Yankee analyst.

"Likely happen" in 2007? That's a far cry from the litany of excitement we've been force-fed over the past twelve months. Intel's PR department earlier this year called the technology "the most important thing since the Internet itself." The Potential for success in 2007 is also a few yards away from saying the technology is "gaining in popularity" as a DSL & Cable competitor; but that's the kind of year WiMax is having: ample sauce, little actual meat.

Intel shoveled out the press releases in 2004, nudging endless outlets to hint WiMax was right around the corner. When said outlets weren't confusing the technology's timeline, they were busy confusing Wimax with other wireless technologies like Flash OFDM, or suggesting it would be a competitor for 3G or Wi-Fi.

Yet "We try not to talk too much about [Wimax]" said Intel's Sean Maloney with a straight face a few weeks ago, his company's 2004 Wimax hype making low-carb marketing seem subdued. If the company had applied the same PR effort to politics as they have Wimax, they could have elected a busted Mickey Mouse lamp to the office of President.

What WiMax is: a promising wireless metropolitan area network (MAN) technology, with a range of 30 miles. It will start as a non-mobile backhaul and T1 replacement technology, possibly evolving into a more promising mobile cable & DSL competitor - if a number of complicated factors (regulation, spectrum availability, competition, other technologies, their own feet) stay out of Intel's way.

What WiMax is not: A miracle broadband solution that in January of 2005 will suddenly begin spewing megabits to desperate housewives in remote corners of North America for pennies a day. Intel only just shipped its first WiMax chip (dubbed Rosedale) in September for trials in 2005. Some analysts believe it could be 2009 before the mobile version of the technology takes serious steps forward in the United States.

WiMax's promise as a backhaul technology is not disputed, but its eventual ability to compete with landline alternatives is. Joseph Crupi, vice president of Texas Instruments' Broadband Communications Group, recently compared WiMax to similar promises made by the Broadband Wireless Internet Forum (BWIF), who pushed a technology known as Vector OFDM that never materialized.

Some of you may also remember HomeRF. Stuck between Bluetooth and Wi-Fi, Intel poured money into that technology until they eventually abandoned ship to join the Wi-Fi party. Right now, while very different technologies, there is not a lot of solid information to suggest that Wimax is any more of a sure thing than HomeRF was.