PALM SPRINGS, Calif. -- Hundreds of engineers, programmers and financial analysts have invaded this community of palm trees and putting greens for a look at good old-fashioned power, Intel (INTC) - Get Report-style. It's time for the twice-yearly Intel Developer's Forum, which the world's biggest chipmaker uses to provide a glimpse of the electronic future.
As it has done at this three-day event in the past, the company will let the world take a peek at its upcoming chips for cell phones, handheld computers, telecommunications networks, cable modems and set-top boxes.
In a key highlight, two long-awaited chips will be unveiled: The Williamette desktops and the Itanium for servers, two products that can propel Intel ahead of its competition in the most profitable areas of the microprocessor market. But the buzz among financial analysts surrounds the possibility that Chairman Andrew Grove will demonstrate for the first time Tuesday a chip that runs at 1.5 gigahertz. If chips were auto engines, we'd be talking metaphorically about a car zipping along at 200 miles an hour.
"They will try very, very hard to put themselves back in front in the speed wars," says Joseph Osha, an analyst at
who has a near-term accumulate on the stock. His firm doesn't underwrite for Intel.
Along with showing its muscle, Intel seems to be using the event to convey another message to the marketplace: Baby, we're back.
Not that the world's biggest chipmaker has gone anywhere. Its stock closed Friday at 105 7/8, just 4 3/16 shy of the all-time high of 110 1/16 it hit Feb. 8.
But the last three years were filled with missteps. In 1997, Intel denied the existence of a market for budget computers that soon exploded in growth. Its first low-end product, the Celeron, was laughed off the market for the lack of on-chip memory (translation: lousy product).
In 1998, Intel's i740 graphics chip for desktop computers was scorched by the competition. The company withdrew from the market a year later. Intel announced a six-month delay in what was then known as the Merced but is now the Itanium chip, a postponement that has now stretched out to a year.
Last year, Intel was
boycotted by privacy advocates for putting a tracking feature in the Pentium 3. Further, its Rambus-based chipset was repeatedly
delayed. The company reluctantly
gave in to pressure from the memory industry to come out with alternative designs.
The slip-ups were so serious that on the company's third-quarter conference call, some fund managers boldly asked for assurance that the company's manufacturing problems were over. A question like that is common from investors in scrappy rival
, but not from Intel investors, who tend to heap praise on Intel execs.
As the forum opens, Intel finds itself in a competitive run for its megahertz in each of its market segments. AMD is pushing into Intel's sacred territory of high-end consumer and corporate desktops;
is sampling a new chip for the server market called the Power 4 that puts two chips on one piece of silicon, which means more performance in a smaller package at lower cost; start-up
is sampling low-cost chips for the notebook market that will work on a fraction of the battery power that a similar Intel chip would use; and
of Taiwan will launch this month a chip for budget computers expected to be priced much lower than Intel's Celeron chips.
And in the new markets Intel is trying to enter -- Web hosting and server farms, cable modems and networking -- it is up against powerhouses
Credit Suisse First Boston
analyst Charlie Glavin says he expects Intel to try to prove that not only are its products better than those of its competitors, but that it can produce them more efficiently. "The big theme will be better control of market segments, lots of contingencies and an extremely
impressive product road map," Glavin says.
Instead of news about delays, Intel's audience will hear of products that are ahead of schedule. And Intel will emphasize a long list of product choices, a turnabout from efforts to dictate the market with singular chip strategies. "Intel used to have a bullet strategy: One size fits all. But the world is going towards more application-specific devices," Glavin says. While Intel seemed to be stretched thin over the past three years trying to make products for a more stratified market, transition time is over, he says.
And, of course, there's speed and power, known in this business as megahertz. On Feb. 7, AMD beat Intel by demonstrating the first 1.1-gigahertz microprocessor at the
International Solid-State Circuits Conference
at the San Francisco Marriott. Behind the scenes, though, there's more than just a macho fight between two fierce rivals, Glavin says. The big sales growth won't be in PCs, Glavin says, but in devices capable of transmitting video images, voice and music over the Internet. To sell in this emerging market, chips will have to be fast.
And they'll need to be out on schedule. If not, then while the desert sun might shine over this conference, there will be cloudy days in Intel's future.