Skip to main content



(AMD) - Get Advanced Micro Devices, Inc. Report

fans. Your favorite company has been getting a free ride lately, but the low-hanging fruit has been plucked.

Look for


(INTC) - Get Intel Corporation Report

to awaken from its slumber within the next couple of quarters, and then we'll see how tough AMD CEO Hector Ruiz and his company really are.

Now that you're ticked off, lets stipulate that Advanced Micro Devices has executed brilliantly and come up with world-class designs. And Intel hasn't. Simply put, AMD deserves its success and Intel deserves its bloody nose.

Nonetheless, the impressive gains in market share and prestige of the last year were made possible by mistakes Intel made a few years ago.Among the worst was a decision to cut capital spending, which in turn was probably brought about by underestimating the strength of the PC market.

The result: A serious shortage of capacity, particularly for chipsets, low-margin products that play the vital role of connecting microprocessors to the rest of memory and other parts of a PC's circuitry. Without them, PCs can't be built. In effect, Intel left business on the table that AMD was only too happy to snap up.

"It's easy for No. 2 to gain share when No. 1 can't deliver enough products. It will get more interesting in 2006 as Intel brings new capacity on line," says Dan Niles, CEO of Neuberger Berman Technology Management. Neuberger Berman currently has a position in AMD, but not in Intel.

Intel's capital expenditures went from $7.3 billion in 2001 to $3.6 billion in 2003 and $3.8 billion in 2004. Last year, spending jumped to $5.8 billion and will climb to $6.9 billion in 2006, the company announced during its

disappointing earnings report this week.

Before going further, it's important to note that guiding a semiconductor company is a lot like steering an oil tanker. It takes a long time to change course. Building a fab (industry speak for chip factory) takes a year and several billion dollars and another year to ramp production to full steam.

Meanwhile, Intel made another mistake. Rather then strengthening its alliance with chipset vendors that were filling part of its needs, Intel decided to compete with them by grabbing market share. And that, says Dean McCarron, principal analyst of Mercury Research, drove those vendors into the arms of AMD.

TheStreet Recommends

That was fine when Intel could make enough chipsets to meet demand, but when it couldn't, Intel had no place to turn because


and other companies were busy producing for the competition.

Making the story a bit more complex is the nature of semiconductor production. High-margin products, like Pentiums, are built on the latest and greatest production lines. When those lines are upgraded, lower-margin products like chipsets move in to fill the void. No chipmaker wants expensive capacity to remain unused for very long.

Intel has been moving its CPU production to a new process called 65- nanometer from 90-nanometer. But for a variety of reasons the transition took longer than expected. And that meant the company couldn't move the CPUs out of the way to make room for the chipsets, McCarron said. And thus was born the shortage.

Although the amount of capacity coming on line this year isn't enormous, it should be enough to ease the shortage. And the big increases of capital spending in 2005 and 2006 should ensure that the squeeze won't be repeated in 2007 or 2008.

Has Intel made other mistakes? Of course. A company that once held the unchallenged technological lead in the industry is now behind AMD in a number of critical markets, such as high-end servers. And those mistakes will not be corrected so soon.

Early Innings

Bad as the week was, with disappointing earnings or soft guidance from a long parade of tech stalwarts, including Intel,


(AAPL) - Get Apple Inc. Report

, and


TICKER TYPE="EQUITY" SYMBOL="YHOO" EXCHANGE="Nasdaq" PRIMARY="NO"/>, and Friday's subsequent drubbing of the


, there are plenty of bulls saying "overreaction."

Most importantly, it's early in the season -- in fact just 20% of the

S&P 500

has reported and just 25% of those failed to make their numbers, noted Eric Thorne of Bryn Mawr Trust Wealth Management. "The market reacted to the high-profile names, but over overall, overreacted," he said during an interview on



Similarly, Neuberger Berman's Niles, said "You still have compelling

low valuations and fairly depressed expectations." But he quickly added: "Don't paint the world with one brush."

And that's exactly what another fund manager, whose company manages $5 billion in assets and requires that he remain anonymous, said: "The market feels schizophrenic. It's hard to justify buying technology in general. You have to look at specific names."

The manager, who is very bullish on consumer spending, pointed to Seagate's

(STX) - Get Seagate Technology Holdings PLC Report

strong report this week as further evidence that demand for computers and related devices is still very strong.

Still, a new report by Gartner, a market research firm, indicates that the PC market won't sustain 2005's torrid 15% growth rate; it will slow to about 10% on a unit basis this year and average selling prices will continue to drop, said analyst Charles Smulders.

At bat next week:


(MSFT) - Get Microsoft Corporation Report




, followed by


(GOOG) - Get Alphabet Inc. Class C Report



(AMZN) - Get, Inc. Report

in late January/early February.