In last month's column, I profiled asynchronous chip design startup Fulcrum Microsystems. This month, I turn to XtremeSpectrum. This startup is in the emerging market for ultrawideband (UWB) silicon and will face the fight of its life in the coming weeks and months.

Founded in November 1998, XtremeSpectrum has bet its 65 employees and $42 million (so far) on a particular physical layer flavor of UWB known as direct-sequence CDMA. Together with ParthusCeva and Motorola, one of XtremeSpectrum's investors, the company is in a hotly contested battle over the IEEE 802.15.3a standard. In the other camp is a group of companies known as the MultiBand OFDM Alliance advocating multi-band OFDM physical layer technology. This camp is formidable and includes Intel, Texas Instruments (also, ironically, an investor in XSI, along with Cisco), other UWB silicon startups and broad-line semiconductor vendors, and a spate of consumer electronics companies.

Before going further, consider what's exciting about UWB. The technology meets the need in consumer electronics for a wireless technology with the bandwidth and quality-of-service to transfer digital video.

Yes, there are opportunities for UWB in applications ranging from asset-tracking to automotive collision avoidance, but the sweet spot sits clearly in the consumer electronics market. Here's why. Think of a consumer with a wall-mounted large-format digital TV such as a plasma TV or an LCD TV. The home theater equipment rack (TIVO, A/V receiver, DVD player, cable STB, etc.) may be located a few feet away or across the room. It would be great to get video from the DVD player to the TV without wires. UWB is the solution, since it combines solid quality-of-service with a minimum bandwidth of 110 Mbits/second at 10 meters, thus exceeding the mid-tens of Mbits-per-second required for HDTV streaming.

You have no business being a startup unless you take a large, though calculated, risk. XtremeSpectrum has taken one in adhering to direct-sequence CDMA for its physical layer. Consumers will win regardless of the PHY technology chosen for 802.15.3a. If multi-band OFDM wins, the UWB silicon market instantly will look a lot like the 802.11 market in terms of profitability, or lack thereof. If direct-sequence wins, XtremeSpectrum wins big.

http://www.eet.com