FWIW
Perpetual Motion Lives
By Rich Karlgaard
Andy Grove giveth, Bill Gates taketh away. That was the joke heard at elite conferences in the PC industry during the 1990s, when ideas such as perpetual motion and a 10,000 Nasdaq seemed altogether possible.
ADVERTISEMENT
The quip happened to be true. Every time Intel 's (NasdaqNM:INTC - news) Andy made a new, smoking-fast chip, Microsoft 's (NasdaqNM:MSFT - news) Bill would follow. He'd mop up the additional cycles with feature-rich software. So Intel would get busy and double the ante again--every 18 months, on schedule, which guaranteed PowerPoint would clean the plate of MIPS within a year. And so it went.
To conspiracy theorists--you Mac users know who you are--it looked rigged. But to the rest of us, it was a miracle of innovation, cooperation, new paradigms and market caps the size of Brazil. It sure looked like a perpetual motion machine.
But now it's 2002. The hangover drags on. Besides, it's no longer Andy and Bill's show. Craig Barrett runs Intel, and Steve Ballmer runs Microsoft. Craig giveth, Steve taketh away? Not so funny. Worse, it may not even be true. Consumers and corporate IT departments are revolting. They don't want more features in Word. Buyers are refusing to shell out money for a new PC every three years. Consequently, Microsoft and Intel have ceased to be fast-growth companies. Don't get me wrong: Both are terrific enterprises with rock-solid financials and decades of future vitality. But that old two-stroke, perpetual motion machine (repeat: MIPS 'n' software, MIPS 'n' software) isn't firing like it used to--and may not again.The New Machine
Fear not. There is a new perpetual motion machine. It is composed not of two parts, but three. They are sensory digitization, search and storage. Think of them as the three Ss. But if you like your trends captured as companies, think Foveon, Google and Extreme Networks.
Digital cameras have been strong sellers for the past two years, but Foveon changes the game completely (see March 18 Digital Rules). Foveon's color sensory chips provide such a leap in color and clarity that their use is sure to become mainstream, even for professional portraits and Hollywood movies. Foveon's shock value is hard to overstate. It's as if some Tutsi tribesmen turned up at Yankee Stadium one day, nonchalantly throwing 140mph fastballs down the pike. That would change everything.
Because Foveon-based cameras are that good, few will want to use film anymore. But here's the kicker. Foveon images are detailed and intricate. They take up enormous amounts of digital bytes--a megabyte for a single photo, compressed, is typical. Yet the photos are so good that people will want them. Hello--demand for storage! We want storage! More storage! More yet!
But every time the storage industry giveth, Foveon is sure to taketh away. Just when we finally think we have enough storage to store all our Foveon photos, along will come Foveon movie cameras! Every man and woman a Spielberg! More storage, please! More! More! More!My Google
Enter Google, everybody's favorite search engine. Google started life in 1998 as a network storage company, of sorts. At least that's what its twenty-something founders, Larry Page and Sergey Brin, had in mind. They ran up their combined credit cards to a max of $20,000, buying cheap, bankruptcy-sale disk drives. (Off the backs of trucks, Brin says.) The two Stanford grad students bought a terabyte's worth. Their goal: to store everything on the Web.
Good search capability, such as Google's, takes plenty of storage. That's fine. Like chips, storage rides its own Moore's Law rate of spectacular improvement. Yesterday's measure of infinity, a terabyte (trillion bytes), becomes tomorrow's cheap box purchased at Fry's Electronics for $150. Now the miracle: Cheaper storage begets more bright ideas from Google. Yesterday's text search for a New York Times article becomes tomorrow's sound search for the opening three bars of an ad jingle. Or a visual search, say, for six MRIs in the world that might unlock the cure for your cancer. Today's breadthy search engines, such as Google, become tomorrow's deep, specific searches, such as Google intends to provide.
I used to think the next perpetual motion machine would be broadband and Web video. That could still happen someday. But a real problem exists with broadband today. It takes an enormous amount of capital investment, which, because of the telecom collapse, is currently off the table. Worse, broadband is regulated by the FCC. Any industry that's regulated in Washington ceases to be fast-evolving, whatever its potential. Too bad.
Sensory digitization, storage and search suffer no such government interference. They are free to evolve as fast as physics and consumer delight will permit. Which is fast, indeed. Have fun with a Foveon camera or Google's search engine, and you'll agree.
Related Links at Forbes.com |
