|
|
Just weeks after Radio Free Afghanistan went on the air, chipmaker Intel last week disclosed plans for Radio Free Intel. No, it's not a new vehicle for delivering propaganda to or from the Santa Clara chipmaker - the phrase itself is ``just a play on words,'' according to a company spokesman. But the initiative it denotes is a big bet for the company, and if successful it could have a major impact on the way we live and work. As described by Pat Gelsinger, the company's newly anointed chief technical officer, in the closing keynote at last week's Intel Developers Forum, Radio Free Intel is a new plan to begin building into the company's chips ``silicon radios'' for wireless networking. And not just some of its chips, but all of them - every last processor and other product the company builds. Starting from research it funded at several universities, Intel thinks it has figured out a way to make complete radios, including amplifiers, antennas and all other necessary components out of the same kind of silicon it now uses to make Pentiums. If so - and Gelsinger acknowledged it's not a 100-percent certainty yet - such radios could be fully integrated into Intel chips in as little as five years. ``We want to get to the point where the radio is nothing more than the corner of the die,'' he said. In case the all-silicon vision doesn't work out, there's a back- up plan that's almost as neat: assemble the radio out of futuristic minimachines - MEMS, or micro-electro-mechanical systems - on a silicon substrate, then build that device into a common ceramic package with a conventional silicon chip. Either way, these radios would be capable of three different kinds of wireless networking, instantly switching frequencies and protocols to handle each as needed: PAN (personal-area networking), the industry term for connecting devices about your person or around your desk - linking a handheld organizer or digital camera to a mobile phone, say, or a PC to a printer or portable music player. That's a role played now mostly by wired USB connections, and one Bluetooth hopes to take over. But by the time integrated radios are ready, Gelsinger suggested, the wireless technology of choice for such applications could be a promising new type of wireless known as ultra-wideband, which received initial, limited approval from the Federal Communications Commission just last month. LAN (local-area networking), for connecting PCs around the office or home to each other and thence to the Internet. In this area several wireless technologies have jockeyed for position in recent years, but one called Wi-Fi, or 802.11, has emerged as the clear winner, and throughout last week's conference Intel officials endorsed it with enthusiasm. Specifically, they are promoting dual-band 802.11 radios, which would implement both the current 802.11b standard and 802.11a, a much faster alternative that's just now appearing in products. (In fact, even before it can deliver integrated silicon radios, Intel hopes to persuade PC-makers to build dual-mode 802.11 radios into every PC, not just notebooks but eventually also desktop models, beginning next year.) WAN (wide-area networking), or connecting to the Internet with no wired connection at all. At this level, Gelsinger said, Intel's radios will be designed to communicate with the cellular networks, which by then will presumably have advanced to full 3G, if not 4G, technologies. In Gelsinger's vision, all these connections would be always on, and they would be automatically updated as you roam from place to place. If Intel succeeds in building radios that occupy just a corner of a chip, the incremental cost of adding this whole range of communications capabilities to digital devices will be nearly nil - that's where the ``free'' comes in. That means not only computers and peripherals, but also consumer electronics gear - TVs, stereos and all the rest. And beyond those devices, Gelsinger suggested some exotic new applications for the technology - baby blankets that monitor an infant's breathing and temperature and automatically send an alert in case of any abnormality; motors that report when they are wearing out; swimming pools that announce when something has fallen in; even ``smart farms'' where sensors wirelessly report on the growth and nutrition needs of each plant. Technology aside, such scenarios raise a host of social issues - about security, privacy, spectrum management (will every product with a chip need FCC approval?), and, by no means least, health (who knows what effect this jumble of new radio signals might have). Gelsinger and his colleagues admit they don't have answers to these questions. But if the technology moves as quickly as they predict, we as a society are going to have to confront them in a hurry. New York Times Syndicate |