February 25, 1999 Web posted at: 12:56 p.m. EST (1756 GMT)
CNN Interactive

In this story: Intel loses the market for sub-$1,000 PCs

Trying to divide and conquer

Is the PC Dying?

RELATED STORIES, SITES

 By CHARLES PILLER

(LA Times) -- ``Intel Inside'' is more than ad copy. Intel Corp. so thoroughly
dominates the microprocessor industry that the slogan has become a defining
reality of personal computing.

Even amid a worldwide semiconductor slump last year, Intel collected a record
$26.3 billion in revenue and earned $6.1 billion as its stock soared to record
highs.

So why should such a steamroller live by Chairman Andrew Grove's creed, ``Only
the paranoid survive?'' Perhaps because for the first time in more than a decade,
Intel faces a confluence of threats that some experts say could number its days
of unchallenged supremacy.

``It's not only a danger for Intel,'' said Nathan Brookwood, an analyst with
Insight 64 in Saratoga, Calif. ``It's in the cards.''

Several factors are disrupting the microprocessor marketplace and casting doubts
on Intel's established practices and long-term strategies:

Federal Trade Commission regulators have sued Intel for allegedly using coercive
tactics to control intellectual property that is key to microprocessor innovation
and to preserve and extend a microprocessor monopoly. The suit, which goes to
trial March 9, could end up limiting Intel's pervasive influence over the
direction of future technologies. Some key industry players expect a new world of
networked information appliances, such as hand-held computers, Web TV and car
navigation systems, to eventually dwarf the PC market and shift the focus of
innovation and source of profits. Intel has only a weak presence in that area. In
1998, Intel lost nearly half its share in the burgeoning low-cost PC market --
dropping from 80 percent to 46 percent, according to San Diego analyst group
Infobeads. Plummeting PC prices have squeezed the historically stratospheric
profit margins that fuel Intel's vast research and development and marketing
efforts.

As recently as 1997, Intel collected 85 percent of industry revenues for PC
central processors, according to Cahners InStat Group, a Scottsdale, Ariz.,
research firm.

``That's an unnatural state of affairs in any real industry,'' Brookwood said.
``As the competitors improve their products and their marketing, that's going to
go down.''

And it has -- dropping to 80.3 percent today.

Ironically, the low-cost PC craze made Intel a victim of its own success. For
decades, Intel relentlessly proved Moore's Law -- the industry tenet promulgated
by Intel co-founder Gordon Moore that projects the doubling of processor power
for the same price every 18 to 24 months. As chips became exponentially faster,
they eventually eclipsed the demands of most PC software.

That means ``for the typical user, higher performance means nothing,'' said
Michael Hackworth, chairman of chip maker Cirrus Logic.

Intel loses the market for sub-$1,000 PCs

In a rare but costly stumble, Intel failed to foresee that many users would
prefer low-cost PCs that still handle typical computing tasks with alacrity over
costlier machines based on Intel's fastest and most lucrative processors. When
sub-$1,000 PCs became a dominant trend, Intel lacked a low-cost product.

National Semiconductor's Cyrix division and Advanced Micro Devices stepped in,
supplying the consumer lines of some big Intel customers, including Compaq
Computer, IBM and Packard Bell NEC.

AMD, known for production blunders that have stalled its periodic forays onto
Intel's turf, had captured more than 50 percent of the market for sub-$1,000 PCs
by last June, according to analysts.

Intel responded with its Celeron processor, an uncharacteristically sluggish
performer that was greeted with derision.

Since then, Intel has retaken some lost ground with an improved Celeron. And it
adroitly combined frequent introductions of new products, such as the recent
launch of the Pentium III chip, with quarterly price drops to keep the pressure
on AMD, which has vowed to stay 25 percent below Intel prices for similar
performance. That pressure became evident when AMD recently said it would lose
money in the current quarter despite growing sales.

But AMD is far from dead. Gateway is reportedly considering AMD chips for some
models -- a potentially major defection by a die-hard Intel loyalist.

And AMD's 3DNow graphics technology has been widely supported by game makers.
Intel will certainly catch up once developers create specialized products that
reflect 3-D instructions in the new Pentium III. But AMD is enjoying another
opening -- this time at the lucrative high end of the PC spectrum that gamers
favor.

In the long run, industry watchers say, even if Intel's lower-cost competitors
achieve only modest success, the raging price war has become a fixture of the
microprocessor landscape. And that's a war on Intel's bottom line.

Trying to divide and conquer

In a counterattack, Intel last year launched a ``segmentation'' strategy. As his
first major initiative, Chief Executive Craig Barrett broke the Pentium processor
family into distinct subgroups: Celeron for economy PCs, Pentium for
high-performance computers and Xeon for server computers. Because server chips
are the most profitable, the Xeon was designed to offset compressed profit
margins from the far larger economy sector.

Segmentation has begun to pay off. Intel continues to dominate the Pentium-class
PC market and has gained ground in lower-cost servers, though it still lags in
the most lucrative segment: powerful servers that cost about $40,000.

``Close to half of our development dollars are going to (servers). Within two
years we will be effectively competing at every price point in the server
space,'' said Patrick Gelsinger, vice president of Intel's desktop products
group. ``We have this view of a billion connected computers. What's on the other
end of that connection? Hundreds of thousands, millions of servers.''

Intel is also making a major commitment to selling networking equipment for homes
and small to medium-sized businesses.

For now Intel might paradoxically derive a benefit from AMD's success. The FTC
alleges that Intel used anti-competitive tactics to acquire or withhold
intellectual property in disputes involving Compaq, Intergraph and Digital
Equipment, now a Compaq unit.

``Our concern with Intel's conduct is that if allowed to continue, roads to
innovation will have to lead to Intel,'' said William Baer, director of the FTC
Bureau of Competition.

But the allegation rests on the assumption that Intel is a monopoly, which the
company hotly denies.

Intel plans to fight that label, citing competitors' surge at the low end of the
market, among other arguments.

``If the FTC is going to be the national nanny and decide who we have to license
to, they are taking away our intellectual property,'' said Peter Detkin, Intel's
assistant general counsel. ``Nobody, even a monopolist, has an obligation to
license their intellectual property.''

If the FTC loses, said attorney Rick Rule, a consultant to Microsoft who headed
antitrust enforcement at the Justice Department under President Reagan, ``then
the courts will have validated the approach to antitrust taken over the last 20
years'': trusting the market to work out problems to benefit consumers.

Rule and others say the FTC holds a weak hand because consumers enjoy continually
declining PC prices while quality improves. But other experts argue that the FTC
is on solid legal ground.

While no one can prove that more competition would have stimulated innovation and
reduced prices, ``antitrust law only requires one to prove that there is a
(plausible) theory of harm,'' said Howard Morse, a former FTC prosecutor who left
the agency shortly before the Intel suit was filed. ``One doesn't have to line up
the dead bodies.''

If the agency ultimately prevails, Intel's ability to set the rules of the road
for how high-tech companies share new technologies would be sharply curtailed at
a time when technological change poses the greatest challenge to the company.

Is the PC Dying?

Intel makes most of its money on chips for PCs, and while no one's predicting the
PC will soon die, many foresee ferocious competition from the nascent world of
Internet-linked digital appliances. Some industry leaders, such as Sun
Microsystems Chief Executive Scott McNealy and Paul Horn, IBM's chief of
research, have already declared the twilight of the PC.

Paul Saffo, a director of the Institute for the Future in Menlo Park, Calif.,
predicts that in 20 years we'll wonder where all the computers went. ``The answer
is they disappeared into the appliances,'' he said.

Intel's Gelsinger isn't buying it. ``The PC is lousy. It's unreliable. It crashes
a lot,'' he said. ``So why are we making 100 million of them a year?'' Intel
won't ignore digital appliances, but it will try to make PCs more appliance-like
in an effort to boost sales. ``We'd like to see a PC on every lateral surface in
business (and) in every home on the planet,'' Gelsinger said.

However, most experts expect the network-appliance industry to grow much faster
than the ``mature'' PC industry. Digital appliances, they say, will ultimately
overtake the PC as a focus of innovation in computing, just as mainframes gave
way to minicomputers, which were in turn displaced by PCs.

Intel has the capacity to take a leading role in the information-appliance age,
said Mitchell Kurtzman, chief executive of Network Computer, a company that
produces technologies for such devices. ``On the other hand, these are the kinds
of discontinuities that create opportunities for others,'' he said.

Intel faces huge challenges in managing those discontinuities. Its dominance in
PC processors means it has been able to steer technology to its own benefit. But
Intel got off to a late start on ``embedded processors'' used in digital
appliances -- licensing the StrongArm embedded processor less than a year ago.

And the embedded market is tougher to dominate because applications vary widely
and strong competitors -- including IBM, Broadcom, Motorola and ARM Holdings --
abound.

So, not surprisingly, Intel is betting that the PC will remain the center of the
computing universe. With a portfolio topping $750 million, Intel is, in effect,
one of the top venture capitalists in Silicon Valley. So it can back up its bets.

The investments tend to stimulate new technologies that require and exploit the
abilities of ever-faster PCs, such as graphics-intensive computing and server
applications that require Intel processors. Other investments push broad-band and
home networking and Intel-based e-commerce tools.

In short, Intel wants to be the company that ties everything together in the age
of the network -- positioning it to profit whichever direction the industry
turns.

The current round of challenges may be daunting, ``but there have been relatively
few periods in the company's history when they haven't had great challenges,''
said Insight 64's Brookwood. ``And to Intel's credit, it tends to do best when it
has great challenges.''