To borrow a line from Kevin Costner’s dreams of an Iowa cornfield: If you build it, they will come. If, that is, you’re a giant chipmaker that can define, create and seed new markets with a single investment.
Building an ecosystem that will support the deployment of fixed wireless broadband access is not an illusory game for Intel Corp. It’s betting a big stack of chips on an emerging wireless broadband standard known as WiMax.
Two years ago Intel incubated an in-house group to explore the potential market for the technology. Now it has an entire business unit behind it, chairs a forum that’s developing and cheerleading industry standards, and its investment arm is pumping millions of dollars into WiMax startups like Motia Inc. of Stamford, Conn.
Intel’s efforts appear to be paying off. Venture investors have seeded at least 11 WiMax startups with roughly $430 million, according to Thomson Venture Economics (publisher of VCJ). Besides Intel Capital, Accel Partners, Band of Angels and Granite Ventures are among the most active VCs in the space, with two investments apiece (see chart, next page).
WiMax is a last-mile network technology that wirelessly delivers broadband into the home. Also known as 802.16-compliant technology, a WiMax system is a wireless alternative to wired broadband connections that operates along a band of unlicensed spectrum. It’s an alternative to cable or DSL that, according to the most optimistic reports, can deliver as much as 70 Mbytes per second across a distance of 30 miles.
A cable company, for example, could build a WiMax network to cover an entire city neighborhood, so that there’s no need to send a truck and a serviceman to flip a switch in every home that orders high-speed Internet. More likely, though, cable companies and telephone companies will use it provide broadband access to rural and remote areas not covered by DSL or cable lines. Cell phone companies that offer email and Internet access on hand-held devices can use WiMax networks to offload some of the traffic on their cellular networks.
Like WiFi, WiMax promises to deliver broadband wirelessly, but the similarities stop there. Technical differences aside, WiMax networks are big enough to cover an entire city neighborhood, while a WiFi hotspot is limited by the four walls of a coffee shop or an airport lounge. A WiMax network is ubiquitous, at least in theory. Residents of a Swedish coastal town called Kungsbacka log on to the Internet with a WiMax system. So do customers of a wireless Internet system in Bangladesh.
A carrier can deploy a WiMax system the same way cell phone companies do, with bay stations and antennas. It would set up a WiMax box on top of an office building with all the requisite antennas and amplifiers, hook it up to a metro-scale network, and use WiMax-enabled chipsets to deliver data at broadband speeds to laptops and desktops.
Intel, which started exploring the potential for WiMax two years ago, is developing a WiMax chipset for carriers that it hopes to ship by Q2 2005. WiMax-enabled laptops should be available by 2007, if all goes according to Intel’s plans. If Intel can put a WiMax chip in every laptop it builds, then any carrier that doesn’t offer WiMax service will be at a disadvantage.
Imagine: If a WiMax icon appears on the screen of every laptop manufactured after 2007, the question for every broadband user will be, which WiMax network do I log on to? Will I get a better deal from a cellular carrier like Verizon or a cable operator like Comcast?
Intel has already signed partnerships with two publicly traded companies, Tel Aviv-based Alvarion Ltd. and Airspan Networks of Boca Raton, Fla., and venture-backed Aperto Networks of Milpitas, Calif. Together, the companies control 40% of the broadband wireless access market, according to Maravedis Inc., a wireless industry research group in Montreal. Alcatel, Fujitsu and Siemens also back the technology.
Since Intel is betting big on WiMax chips for computing devices, the opportunity for WiMax startups is in hardware, software and services. That’s where investors are laying their bets. They’ve financed companies like Aperto Networks of Milpitas, Calif., which builds wireless broadband systems for carriers as far away as Sweden and Bangladesh; Orthogon Systems, a broadband access provider based in Ashburton, U.K.; Motia, a Stamford, Conn.-based startups that makes specialized chips to operate “smart antennas” for wireless providers; and WiDeFi, an 18-month-old startup in Satellite, Fla., that makes technology to boost the performance of wireless networks.
Maravedis expects the market for WiMax technology to reach $600 million by 2005 and grow at a steady clip to reach $1.6 billion by 2008, including chips.
The challenge facing investors-Intel included-is that there is no guarantee that the market will ever come to life. There’s uncertainty over whether the incumbent carriers will buy into the technology and, if they do, to what extent. AT&T has joined the WiMax Industry Forum (a forum headed by Intel executives), though it’s made no major moves in the sector, while British Telecom has several field trials in progress in the United Kingdom. While research commissioned by Intel predicts that WiMax will reach 7 million subscribers worldwide by 2007, IDC forecasts that there will be 124 million DSL lines in service that same year.
Then there are technical hurdles. WiMax operates along unregulated frequencies, making flexibility inherent within the standard. There’s no guarantee that one piece of equipment will talk to another. And as WiMax communications leap across frequencies, they may interfere with others, prompting the FCC to step in and slow the sector’s development, as it did with ultrawideband technology.
Betting on WiMax is betting on the fact that a single company can create and define a new market. As one wireless industry analyst puts it: “If Intel wasn’t involved, the whole thing would go away.”
Intel has done it before. When the semiconductor industry was still using 180-nanometer chips, Intel announced that its next milestone was the 90-nanometer chip. Work on 157-nanometer chips came to a virtual standstill. Plus, Intel’s Centrino chipset was the first one developed specifically for wireless Internet access. It’s now ubiquitous in all laptops built after 2003. Not only was WiFi an in-house research and development project at Intel, but Intel Capital seeded a handful of companies to manage and deploy wireless broadband networks-so many that you can log onto a hotspot at McDonald’s while you munch on your Filet-O-Fish.
If history repeats, WiMax will be more than a field of dreams.