OT - Interesting article:
Changes mean end of status quo
Intel's fantastic run could be finished.
Over the past five years, Intel has become the absolute dominant force
in personal computer hardware, dictating the very pace of technological
change. Thanks to that clout, the company has grown from $8.8 billion
in revenue in 1993 to $25 billion in 1997. Its market value of $138
billion exceeds that of Chrysler and Ford combined.
Yet personal computing is changing into something that might be called
everywhere computing. The one company most likely to get hammered
by that shift is Intel, maker of microprocessors in 85% of PCs.
The trend is long term and has little to do with Intel's difficult 1998,
which has been marked by earnings declines, job cuts, temporary plant
closings, a government antitrust action and the delay of Merced, an
important future chip. And its stock is up because Wall Street thinks
Intel's business of making microprocessors for PCs has always been
cyclical - and the cycle is about to turn up.
But what if it doesn't? Increasingly, industry watchers are expecting
Intel to bounce back a bit, then get stuck in low gear, growing a few
percentage points a year. Last week, venture capitalist Andy
Rappaport of August Capital gave a talk titled: "The computer market
in the post-Intel era." Says technology investor Roger McNamee: "The
Intel question is: Is this a transition or is it over?"
Over wouldn't mean the demise of Intel, just its fading as a booming,
unbeatable power in computing - like the point when the British Empire
began to ebb. And if Intel's reign is done, the reason would be that the
PC, Intel's power base, is getting knocked from its perch.
The focus of computing is moving to the Internet and networks and
away from the computers themselves. As that happens, more kinds of
devices - PalmPilots, WebTV, cell phones, network computers - can
attach to networks and take over some of the tasks that used to belong
only to the PC. That makes a powerful PC less important and a
powerful network more important, which is like an earthquake under
Intel's feet.
"The wall has been broken that says you need the latest Intel chip and
PC and Windows to get the latest applications," says Ed Iacobucci,
chairman of Citrix Systems, which is both an Intel partner and a maker
of networking software that could damage Intel. "It gives new degrees
of freedom. A lot of people don't like that. It was pretty cozy before.
Now there's this intrusion, and our industry is just going to have to deal
with it."
Intel chief had a plan
Intel has seen some of this coming. For years, Andy Grove, Intel's
chairman, has had a strategy to make the PC "it." He's wanted to make
the PC the machine we use for everything: working at our desks,
playing games, communicating via PC-based video phones, watching
TVs run by a PC-based set-top box, and so on. Since Intel is already
in the vast majority of PCs, it can't really grow market share - it has to
grow the market.
"So we do a lot of things to create new users and new uses for PCs,"
says Intel CEO Craig Barrett. "And when you see us supporting a
music festival on the Internet or investing in small companies that have
great ideas for PCs, all of that is designed to grow the pie. We do a lot
of evangelizing."
Up to this year, the strategy has worked. PCs have pushed into more
homes and offices, becoming an important communications device and
the way many people play video games.
At the same time, Intel has consistently made its chips - and thus PCs -
faster and more powerful, drawing many buyers into a cycle of frequent
upgrading.
"You want people to have more processing power, a bigger hard drive,
more D-RAM (dynamic random access memory, a computer's
working-level memory), a better user experience associated with that,"
Barrett says.
Intel's partner in the strategy has been Microsoft. Intel would build a
better chip, and Microsoft would create a new version of Windows and
other software that would have to run on that chip. The two are so
wrapped together and so dominant in PCs, they're often called Wintel.
But to be high-growth and high-profit, the Wintel strategy is dependent
on the constant-upgrade cycle. And that cycle is being chipped apart.
Business goes networking
In businesses, the PC is being challenged by things such as lower-cost
network computers (NCs) and by powerful networks that let workers
do their jobs on aging PCs.
No question that inside corporations, Wintel PCs are alive and well.
"PCs, cheap or otherwise, aren't disappearing anytime soon," says Kim
Polese, CEO of Internet company Marimba. But the NC is plowing in.
An NC is a lean PC, with a less-powerful, lower-priced chip and used
mainly to access networks. NCs often run software based on the Java
language.
For some businesses, a PC's power is overkill and not worth the cost.
Prime Equipment, a Houston company that rents out construction
equipment, has replaced 90% of its PCs with IBM Network Station
NCs. IBM says it has 3,000 customers worldwide using its NCs. "We
expect to see increasing momentum for NCs," says Irving
Wladawsky-Berger, IBM's Internet chief.
Many times, NCs replace old dumb terminals that were attached to
mainframes in places such as airline ticket counters and retail outlets.
Still, a year ago, those would probably have been replaced by Wintel
PCs. Some of IBM's NCs will be powered by Intel chips, but ones not
as powerful or profitable as Intel's PC chips.
Citrix is potentially more damaging to Intel than NCs. Its ICA software
allows old PCs, attached to a network, to run the latest version of
Windows and other applications. The software resides on the network
on a powerful computer called a server. To keep up to date, the
customer has to upgrade one server instead of many PCs. New
versions of ICA allow devices such as a PalmPilot to run on Windows.
At least 47 companies are building ICA into devices. "It makes what's
on the desktop irrelevant," Iacobucci says. "That's bent a few brains a
little bit."
Citrix also might be the chain saw that cuts Wintel apart. Microsoft has
licensed ICA and is building it into Windows NT - Windows for
networks - meaning it's making sure it can still grow in a non-Intel Citrix
environment.
Intel hurt at home, too
Intel's not getting a break in homes, either. Consumers now want PCs
to be dirt cheap - 40% of the home market is for sub-$1,000 PCs. At
those prices, Intel's decision is to either make very little profit on its
chips or cede the market to other chipmakers, such as Advanced
Micro Devices and National Semiconductor, which are giving Intel its
stiffest competition in a decade. On top of that, interlopers such as
WebTV are stealing consumers who just want to get on the Internet
and send e-mail. If most of what consumers want to do is on the
Internet, they'll see less need to spend a lot on a PC.
In fact, in many consumers' minds, excitement about hardware has
shifted away from PCs and toward funky hand-held devices like the
PalmPilot. Two million will be sold this year, up 50% from 1997. PC
sales this year will be up just 12% over 1997. And a flood of devices is
coming, from phones that can show Web pages to video-game units
that can access e-mail.
Intel's role in the device market is a question mark. It has been unable,
for instance, to win a position as a maker of cable-TV set-top boxes.
"There's a collision coming in the device space," says Bud Tribble, an
executive at Sun Microsystems, an Intel competitor. "There's not a
common chip or operating system, no architecture with critical mass.
It's a green field."
Intel, though, can't be counted out. It's a company of tremendous
strengths in technology, manufacturing and management. It controls the
high-end desktop computer market, which will grow and be profitable
for the foreseeable future, and is gaining clout in servers and even
supercomputers behind new chips like its XEON. Intel has $10 billion
in cash, which it could spend to buy technology or a company to get
into chips for devices. And Intel is not blind to the shifts: Grove has said
Intel must move into devices and low-cost chips. In fact, the company
is aggressively pushing its low-end Celeron chip. "Celeron is one of our
fastest-ramping products ever," says Intel executive Paul Otellini.
Intel also has some time. The shift away from PCs can't get into full
swing until networks get better. "The networks are too slow to
download all the applications and all the data" people want to get on
devices, says James Cannavino, chairman of SoftWorks software
company.
At any rate, Intel has never before shrunk from a battle and no one
expects it to this time.
usatoday.com
|
