Last week, the wireless chip startup WiQuest shut its doors after its VC backers gave up on the maker of ultra-wideband (UWB) wireless chips. The closure of the firm with 120 employees and $54 million in funding raises a question about the nascent UWB technology, which provides high-speed data wireless links over a short distance. But supporters of the technology say one company’s failure won’t sink the industry.
For years, UWB technology has been a solution looking for a market. In essence, it’s meant to move high-quality digital video from one gadget in a room to another at a high speed.
The WiMedia Alliance, which has more than 350 member companies, developed the standard based on the UWB technology. UWB devices can transfer a theoretical 480 megabits a second of data over a couple of meters and about 110 megabits a second over 10 meters. It was a good alternative for transferring high-definition video within a room from a set-top box to a TV. Wireless USB wire replacement equipment can transfer data at about 220 megabits a second over a few feet.
The technology itself can be used for a variety of applications. That flexibility prompted one group of companies to use UWB as the technology for wireless USB (universal serial bus), where the wireless technology could replace the wires that connect a computer to a monitor or a keyboard to the computer. The WiMedia Alliance also hopes that the Bluetooth standard will adopt UWB in faster versions of the Bluetooth wireless connectivity solution, which is used as a “personal area network,” or PAN, for connecting devices to each other or to printers.
But the shutdown of Sequoia-backed WiQuest in Allen, Texas, raises big questions. Intel was also once developing a UWB chip set but abandoned that project in favor of investing in UWB startups. Linley Gwennap, an analyst at the Linley Group, says you can probably stick a fork in this technology now.
One of the problems was that the companies that led with the technology tried to get other companies to back their version of a standard. But a standard battle erupted that took several years to resolve with Intel, Texas Instruments and Staccato on one side and Freescale and others on the other. The WiMedia standard was only adopted at the end of 2005.
The first chips came out two years ago, but they were expensive solutions requiring two or three chips. And in reality, those chip sets transmitted data far below the theoretical speeds. Market researcher In-stat said only 100,000 or so UWB chip sets shipped in 2007, with the figure expected to grow to 190 million by 2012. Products that have hit the market include a wireless USB adapter from Kensington.
The slower-than-promised chips were a disappointing outcome for a technology that had been promised for more than a decade. The Federal Communications Commission started looking at proposals for the use of air waves for UWB technology in 1998. It finalized its ruling in 2002.
It allowed companies to use a broad frequency range of 3.1 gigahertz to 10.7 gigahertz of the wireless spectrum, but it limited the range to 10 meters or so in order to ensure there was no interference with police, fire or military radios. The first low-frequency chips did run into interference, so now the new chips are focusing on 6-gigahertz and up.
Sponsored by VB
There are now single-chip solutions coming out, but WiQuest’s own single chip solution came back from the factory dead on arrival. That is, the first prototype didn’t work and it meant the company would need months of reworking. Rather than put more money into the company to fix the chip, the VCs shut it down. Stephen Wood, the WiMedia advocate at Intel and president of the WiMedia Alliance, still believes that one company’s failure won’t doom the industry. He believes that cheaper chips will solve the adoption problem.
“This is a natural phase we are going through where the technology is evolving, the industry is maturing, and consolidation is happening,” Wood said.
The first chip sets were $15. At $8 a chip, where many of the solutions are now, laptop makers and desktop computer makers aren’t likely to adopt the technology. Once it moves to $4 a chip, then Wood believes adoption will take off, just as it did with technologies such as Bluetooth. But a technical article notes that prices are at about $20 now and will hit $5 in 2010.
Wood noted that the technology started with 14 startups, a situation that everyone knew wouldn’t last. Today’s surviving UWB startups include Alereon, Wisair, TZero Technologies, and Staccato Communications. Other players include Realtek, NXP/ST Microlectronics, Sigma Designs, and CSR. Samsung demoed wireless USB technology a couple of weeks ago, hooking cameras, cell phones, TVs and computers to each other via Alereon’s chips. Laptops with wireless USB options built into them are coming from Dell, Lenovo, NEC, Fujitsu, and Toshiba.
But some companies are shifting gears. Radiospire is now shifting from UWB to a new technology, dubbed 60-gigahertz wireless, for transferring high-definition video. SiBeam has also moved ahead with its own rival Wireless HD technology for transferring data using the 60-gigahertz spectrum. TV makers such as Sharp, Panasonic, Mitsubishi and Samung are supporting it.
Meanwhile, Wi-Fi is trying to move up from the 54-megabits per second range to several hundred megabits a second via the 802.11n standard. But Wood says Wi-Fi is still fraught with transmission quality problems. He notes that both Bluetooth and Wi-Fi, which are now pervasive in a wide range of gadgets, also had very slow starts. New radios can take five or more years to gain adoption, especially with regulation and the development of new chips involved.
Over time, new uses will emerge, as people start to download movies from set-top boxes or the TV directly to handheld devices such as the iPods of the future. Wood sees the day when you can download a movie from a kiosk to a handheld just before you get on an airplane. You could do the same for gadgets in cars, or upload camcorder videos wirelessly to a TV. That could be a long way away from today, unfortunately.
VB's research team is studying mobile user acquisition... Chime in here, and we’ll share the results.