SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Qualcomm Incorporated (QCOM)
QCOM 174.76+0.3%Dec 23 3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Craig Schilling who started this subject9/24/2000 12:24:04 AM
From: Ruffian   of 152472
 
A View From the Gildercosm
By Tiernan Ray

BY THE TIME Steve Forbes tore into the Democratic Party for its
"fuzzy language" and "fuzzy ideas," his audience at last week's
Gilder/Forbes Telecosm Conference near Lake Tahoe, Calif.,
might have been excused for feeling a little fuzzy themselves.
After all, they'd been pummeled for a half hour beforehand with
bits and bytes about erbium-doped fiber amplifiers and
micromirrors.

Technology uberpundit George Gilder — the hard-charging host of
this $5,000-a-plate event and Forbes's friend — opened his annual
technology confab with a kind of visionary assault that didn't
spare the technology-challenged. Beating on the microphone in
his breathless John Wayne accent, Gilder regaled his audience
with reasons why (to crib from the subtitle to his newest book)
"infinite bandwidth will revolutionize our world."

"Well, today you can put 1,000 wavelengths on a single strand of
fiber," Gilder intoned. "And 864 fibers in a single conduit. And if you run 10 billion bits per second over
each of those wavelengths, what do you have? That's right, one petabyte of information every second!"

Got that?

Doubtless some of the assembled investors were left behind as they munched on minimuffins, not 20
miles from Donner Pass (which was appropriate given one of Gilder's central theses — that the future
cannibalizes the past). But they seemed to enjoy themselves anyway as they floated along on Gilder's
characteristic surge of optimism. Folks go to Gildercosm as much for passion as for technological
insight. And Gilder is a ready font of both.

Reality Bites
But some invitees to the conference spoke of a decreased sense of exhilaration this year. Perhaps it had
to do with the fact that Gilder's astute vision of the "Fibersphere" (his word for the vast array of new
fiber-optic technologies) has finally started to bear fruit. While that's a cause for celebration, it also meant
the panel sessions tended to become mired in prosaic product details and obscure sectarian battles. It
was as if the revolution had finally been completed, and the difficult job of building a real society had
begun.

The problem is, Gilder's vision of limitless bandwidth is a lot more complicated than it looks. His previous
success was a book, "The Meaning of the Microcosm" — the title deriving from the Greek "mikros
kosmos," or "little world." It elegantly described how Intel (INTC) revolutionized computing by shrinking an
entire computer down into the fine lines of a microprocessor etched in silicon.

"Telecosm" is obviously meant to suggest that in a similar way, fiber-optic tricks like dense-wavelength
division multiplexing, or DWDM, create a whole network-within-a-network by multiplying indefinitely the
number of wavelengths in a fiber, and hence the number of communications channels. Unfortunately,
communications isn't a self-contained world like the computer: The entire Internet revolution is dependent
upon the interconnection of different networks, not just multiplying the area inside a fiber. And what the
Gilderites are discovering is that adding constituencies to the process ratchets up complexity very
quickly.

To Infinity and Beyond
Gilder's vision is certainly compelling. The premise of the conference — as explained in the
complimentary signed copies of Gilder's book, "Telecosm" — is that new communications technologies
are unleashing an abundance of bandwidth that will overwhelm the old abundance of computer processing.
Traditional dial-up modems, driven by simple computer chips, are being replaced by optical physics with a
near-limitless ability to move data.

In terms that have made him the Count of Coinage, Gilder
explains that the Fibersphere presages an era of infinite
bandwidth. In this world, rather than charge $14.95 for an
Internet connection, phone companies will waste bandwidth
like online publishers waste words. They will toss a fiber-optic
cable into your living room for free in return for your everlasting
loyalty in buying high-bandwidth products and services.

The event was largely a parade of Gilder's intellectual allies in
the bandwidth wars — the men and women charged with
making this future happen. In the quasispiritual atmosphere of
the Telecosm, Gilder fuses technology with a conservative
politics in which the rugged soldiers of science work around
government regulation with sheer ingenuity and pluck. He lauded heroes like Global Crossing (GBLX)
Chief Executive Leo Hindery, whose undersea fiber-optic empire is challenging government-backed foreign
telecom firms. Gilder sounded like The Duke himself in "The Longest Day," telling the troops, "Don't give
the enemy any chance. Send 'em to hell."

Huber's Law
One of Gilder's favorite optical warriors is Dr. David Huber, founder of fiber-optics giant Corvis (CORV). He
took the rostrum to reflect on a lifelong passion for optical networking. "When I entered school, I had a
choice to get an electronics degree, or to pursue optical physics," he said. "Optics seemed to me to offer
lots of promise, and it's turned out to have been a good decision." Given his $5.6 billion in Corvis stock,
who among the History or English graduates would disagree?

Huber laughed at Moore's Law, the 35-year-old rule of thumb in the computer world that says the
information capacity of semiconductors increases by 100% every 18 months or so. In contrast, Huber
said breakthroughs in DWDM and high-speed lasers had brought about a 16,000-fold increase in the
information-carrying capacity of fiber in the past few years.

"Moore's Law is a snail," sneered Huber.

There were visions of the Fibersphere to come, too. Terabeam, a wireless start-up in which Lucent
Technologies (LU) has invested nearly half a billion dollars, showed off a demo of a broadband access
network that sends beams of light from rooftop transmitters through an office window to a receiver
resembling a very large parking meter. From just over a mile away, the system beamed to the conference
three DVD movies (uncompressed), a high definition television signal and a two-way video conference —
all of it passing over a link that could be ratcheted up instantaneously to one billion bits a second from 10
million bits, based on the subscriber's preference. The system should roll out commercially early next
year.

Simon Cao, a scientist with component vendor Avanex (AVNX), said it would soon be possible to stuff
not 1,000 wavelengths of light into a single fiber, but 100,000 — each one carrying 10 billion bits of data
every second, possibly more. He predicted that increasing sophistication in the manipulation of optical
physics might even lead to entirely new ways of representing information. "It's much more natural for a
photon to have its frequency changed, than to switch by other means," said Cao, somewhat
enigmatically.

Filling the Vacuum
But even in a world of superfibers and infinite bandwidth, things aren't as easy as they seem. For one
thing, there was little discussion of the looming scarcity in the optical world, namely the slow pace of
manufacture that has stymied supplies of components and raw fiber. Nor was there much consensus on
which technologies will best achieve the vision. Theoretical bickering on the panels is a likely harbinger of
confusion up the road.

Another big question: How much bandwidth is enough?
Matthew Bross, the chief technology officer of Williams
Communications (WCG), mounted the podium with a child's
toy in hand — a spiky, colorful plastic ball. "These are all the
different communications channels — fiber, digital modems,
cable modems," said Bross, pointing to the plastic spires
jutting from the ball. "The ball's center represents data traffic,
and it's one-and-a-half inches across. But as you increase the
communications capacity...." He tugged at the spikes, pulling
the corners of the toy apart until it unfolded completely into a
huge geodesic sphere 15 inches in diameter. The crowed whooped with delight. "The traffic also expands,
by an order of magnitude, for a 98,000% increase in the volume of traffic!"

Bross's point: Data, like air, expand to fill a vacuum, and no amount of bandwidth will ever be enough.
(Which makes you wonder what kinds of applications will strain Gilder's petabyte communications
channel down the road.)

Of course, at the end of the day, it's scarcity, not abundance, that leads to profits in the real world, a fact
made blindingly obvious when one of Gilder's heroes, Metromedia Fiber Networks (MFNX) CEO
Nicholas Tanzi, said that while bandwidth might be essentially free, "We have a premium network, and
people will pay a higher price for that." Nayel Shafei, CEO of start-up Enkido and the panel's moderator,
shouted with pleasure, "You see, George, the bits won't really be free!"

The point is, the Fibersphere isn't an island. Avi Freedman, an executive with hosting firm Akamai
(AKAM), raised his hand at one point to ask how Metromedia and the rest could charge premium prices if
they still had to connect their cables to those of nonpremium phone companies owning dumpy old
electronic networks. The lowest common denominator, in other words, would limit the promises of fiber
optics for some time.

A Gigabit in Every Pot
Bouyed by Gilder's enthusiasm, the pioneers at Telecosm want to believe their command of optical
physics will help them surge past the vendors who've sold all the software that makes the Internet run —
companies such as 3Com (COMS), Cisco Systems (CSCO), Juniper Networks (JNPR) and anyone
else wedded to plain old electronics. Gilder himself downplays the interconnected nature of networking,
not because he doesn't anticipate it, but because it doesn't fit with his aesthetic, much the way the need
for government regulation and some measure of bureaucracy doesn't fit with his rugged politics.

A more complex future was offered by Gordon Stitt, CEO of Extreme Networks (EXTR). Trotted out at
one point during the conference as something of a sacrificial old goat, his company makes that boring
electronic equipment whose demise Gilder has foretold. Except that Extreme's gear increasingly is being
bought by phone companies to sell Internet service based on Ethernet, a 28-year-old technology that's the
most popular networking medium in the world.

No, Ethernet couldn't speed up as fast as fiber optics, Stitt admitted, but it's already everywhere. In his
own soft-spoken way, he defended Ethernet as the great leveler, the most democratic of technologies, as
something akin to a chicken in every pot, a gigabit of bandwidth in every home. Stitt suggested the
simpler, less razzle-dazzle technologies that fostered the Internet — and that most companies already
understand — would be around for a long time to come, Telecosm be damned.

As with any revolution, the devil is in the details.
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext