In Depth: Intel's Chip Plans Give WiMax A Mighty Push Forward

We all want the same thing when it comes to a wireless Internet connection: coverage everywhere, superfast speeds, not too pricey.

What we don't know is when we're going to get it. On that front, Intel last week pushed the zoom-ahead button, disclosing plans to deliver by year's end a new chipset called Rosedale 2 that should make it easier to access WiMax from mobile computers. By early next year, Intel predicts some PC makers will be building those chips into laptops, and that may be the jump start the industry needs. Think of the role having Intel's Wi-Fi-friendly Centrino chips in so many laptops played in the proliferation of hot-spots in offices, homes, and public places the past few years.

Intel thinks WiMax can pick up where Wi-Fi leaves off. It promises speeds that outgun cable modems and spans distances of several miles from a base station. The implications for business computing could be far reaching. WiMax could allow new competitors into the Internet access market, letting businesses put a fixed WiMax receiver on their buildings as the big pipe for Internet access service, which they then route to employees using existing Ethernet networks.

WiMax also presents a new option for delivering broadband to remote offices, as telecommunications companies fill gaps in coverage where their DSL or other broadband doesn't reach. Those things are happening on a very limited scale in the United States, with larger cities such as Seattle and Honolulu awaiting implementations of WiMax-related technology soon (see story, "WiMax's Bottom Line"). Outside the United States, the uptake is faster, with several large-scale deployments under way in major cities.

If WiMax becomes a feasible channel to get fast Internet access across wide areas, it would create a take-it-with-you Internet that would spark new business uses and even new devices. Today, mobile Internet access depends on third-generation cellular technology that, though improving, isn't particularly fast or cheap.

Intel alone isn't powerful enough to make WiMax succeed, and there are critical differences that'll make this a tougher slog than Wi-Fi. The need for scarce licensed spectrum is the biggest. Also, most established U.S. telecom companies are lukewarm about WiMax, though BellSouth last week expressed fresh interest in testing it. Countries without established telecom infrastructures are getting a head start on the United States in their use of WiMax. But if Intel can get WiMax-ready chips built into laptops, the potential for success across markets shoots up. "As a foundation inside the company right now, the mobile Internet is mobile WiMax," says Scott Richardson, VP and general manager of Intel's service provider business group.

Chicken-And-Egg Problem

WiMax comes in two flavors: fixed, based on the 802.16d standard ratified in mid-2004 that, while wireless, doesn't allow for much movement away from the wireless router that receives the signal; and mobile, or 802.16e, which is meant to compete alongside 3G cellular technologies for data transmission, letting people communicate while walking or riding in cars.

Despite its runaway Centrino success, Intel hasn't always bet right in the mobile market. Just last week, it bailed out of smart-phone communications chips by selling its business to Marvell Technology for $600 million after investing more than $3 billion. But it's not running from the mobile world. The Rosedale 2 chipsets are designed to support both fixed and more advanced mobile WiMax systems. Intel already sells a WiMax chipset called PRO/Wireless 5116--its first Rosedale chip--but that only supports the fixed version of the technology and comes embedded in bulky modems.

These chipsets will help drive the proliferation of WiMax-enabled devices, predicts Paul Sergeant, director of marketing for Motorola's wi4 WiMax products. Motorola is testing mobile WiMax in Pakistan, where Wateen Telecom, the country's major service provider, in May hired Motorola to design and deploy a nationwide wireless broadband network based on WiMax. In one of the most-ambitious mobile WiMax efforts yet, the company expects to support more than 1 million users, with the initial deployment expected to be completed by year's end.

It's that elusive combination of telcos providing service and people having the technology in hand to receive it that WiMax hasn't achieved. "There's a chicken-and-egg problem where laptop vendors and manufacturers won't start putting WiMax chips into laptops and other devices just because Intel recommends it as a good approach," says Charles Golvin, an analyst at Forrester Research.

Intel intends to spur that change with a chipset that provides the Wi-Fi capabilities people know they want with the WiMax they likely wouldn't pay extra for at this point. The company expects to have an integrated fixed/mobile WiMax and Wi-Fi chipset, code-named Ofer-R, ready by 2008. Intel's vision is embedding chipsets on "ultramobile" PCs with 3-inch screens. Fujitsu pushed the WiMax vision further last week, announcing a mobile WiMax chipset it says will be available in August 2007 aimed at cell phones and consumer devices like cameras.

The Spectrum Question

WiMax could be a great way for well-funded newcomers to compete with incumbent carriers, except for one thing: The licensed spectrum it requires is scarce and expensive in the United States. The WiMax Forum, which certifies products for interoperability, has blessed only two bands: 2.5 GHz, which is mostly owned by Sprint Nextel and ClearWire, and 3.5 GHz, which isn't available in the United States. Sprint said last week it's evaluating several wireless technologies, known as 4G, to create an ecosystem of manufacturers, vendors, and operators to provide new services in the 2.5-GHz band. Sprint hasn't committed to a technology and says it's considering ones based on cellular and WiMax. However, at a Wireless Communications Association presentation last week, Sprint spoke more about Flash OFDM, a contender for mobile broadband, and an outgrowth of the UMTS standards used in Europe than it did about WiMax. Sprint will make a decision this summer.

BellSouth's interest in WiMax is typical of many U.S. service providers: It wants to fill in gaps in rural areas or densely populated urban areas that are hard to reach with DSL. BellSouth owns spectrum in the 2.3-GHz band and is awaiting approval from the WiMax Forum to use it for a WiMax deployment, says Mel Levine, BellSouth's director of wireless product management. Intel plans to have its chipsets work in the 2.3-GHz band as well as 2.5 and 3.5. BellSouth said last week it will begin lab tests next quarter using mobile WiMax technology from Alcatel to find ways to offer wireless broadband Internet access. Commercial services could come as early as next year.

Verizon Communications is testing WiMax equipment it may use for niche apps in places it doesn't offer wireline broadband services. Cingular, T-Mobile, and Verizon Wireless also say they have no plans to deploy U.S. WiMax networks, although T-Mobile International has conducted several trials and recently deployed a service based on the technology on British trains.

More U.S. spectrum could become available. Decades ago the federal government allocated part of the 2.5-GHz spectrum to schools for educational television programming that never took off. Now the schools have to use it or see it auctioned off. Some, including the Milwaukee Public Schools, are evaluating WiMax's potential to provide wireless broadband access in students' homes. Other auctions could yield even more-valuable spectrum. The FCC will sell 1,100 licenses in the 1.7-GHz and 2.1-GHz bands this August. Lower frequencies would let WiMax penetrate homes better and be broadcast over wider areas.

There's A Business Need

An estimated 175 WiMax trials have been launched worldwide, with 35 commercial fixed service offerings already up and running--mostly in the 3.5-GHz band, Intel says. WiMax networks can be found in Croatia, Saudi Arabia, South Africa, and Taiwan, where the appeal of delivering broadband without laying expensive fiber outweighs the risk of working with still-emerging technology.

Some U.S. businesses are turning to vendors like TowerStream that provide WiMax-based broadband as a T1 replacement or backup. TowerStream offers service based on fixed WiMax with VoIP in parts of Boston, Chicago, Los Angeles, New York, San Francisco, and Providence, R.I. ClearWire, the company founded by cellular pioneer Craig McCaw, sells wireless broadband using proprietary technology related to WiMax in a number of midsize markets.

There are other fledgling markets for WiMax in the United States. Just ask Manish Gupta, marketing VP at equipment maker Aperto Networks, who rattles off municipal wireless networks from companies such as EarthLink, wireless coverage for cable operators, and organizations with lots of mobile employees, like police or utilities. The FCC's deregulation of DSL goes into effect in August, meaning phone companies such as AT&T and Verizon no longer must let independent Internet service providers such as AOL and EarthLink run DSL services through their wires. ISPs will need an alternative access pipeline, and WiMax is a candidate. The WiMax Forum predicts cities will become "metro zones" blanketed by wireless broadband, once WiMax is incorporated into mass-market laptops and PDAs.

3G cellular networks also will continue to grow and, with improved technology, deliver better data service and coverage. The major PC makers this year started offering or made plans to offer laptops with EV-DO chips and built-in cellular data modems. WiMax also isn't likely to displace existing wired technology such as DSL and cable. Instead, mobile WiMax likely will work in concert with existing networks to fill in and extend coverage. U.S. businesses in the next year can expect to see fixed WiMax as an increasingly viable option in some areas for what Aperto CEO Michael Pratt calls "competitive bypass," an alternative to conventional telecom companies, at least for certain niche services like backup and remote office connectivity.

IT managers are watching these technologies, but what they really want is to not have to worry about them. Robert Neill, director of IT systems at ATP, which governs men's pro tennis, says what's needed are devices that let people use any available network--Wi-Fi, 3G cellular, or WiMax. "We need some method to persistently gain access to these physical carriers with one logical account, meaning cooperation among carriers similar to the way we're able to use a mobile phone or BlackBerry," Neill says.

As for the dream of videoconferencing over the Web, seamlessly switching across WiMax networks while sailing along on the commuter train or riding in a taxi at 70 mph? In this untested market, that one has a long way to go. But for signs of progress, keep an eye out for those WiMax chips showing up in laptops.

Continue to the sidebar:
WiMax's Bottom Line