SANTA CLARA, Calif. - Intel Corp. is the king of microprocessors, the semiconductors known as the brains of computers. Since the 1980s, the Silicon Valley giant has stayed ahead with a simple mantra: Make its chips faster.
But 2 1 /2 years ago, Intel made a risky decision to abandon that narrow focus. Speed was becoming less important to personal-computer users, who mostly used the machines to surf the Web and run a few simple programs. So, in a shift that transformed its culture, Intel pushed its engineers to pursue an entirely different goal: Build chips that fit the new ways people actually use their computers.
That bet is beginning to pay off, helping Intel become one of the biggest winners to emerge from the long tech slump. In the new world of digital wireless gadgets, computer users care about other things besides speed — such as long battery life and small size. Intel has reorganized the company to deliver chips that offer just that.
Intel recently reported its highest third-quarter revenue growth since 1996, amid a broad comeback in personalcomputer sales.
Net income more than doubled from the same period last year, and profit margins are approaching the record levels set during the tech boom. The company posted all-time sales records in most product categories, in a quarter that is not usually its strongest. Its stock has more than doubled this year to about $32.
Intel got a big boost from one of the first fruits of its new strategy: The Pentium M, a specially designed microprocessor for laptop computers. It’s no speed demon, but it gobbles a lot less power than typical chips, giving portable computers a couple of extra hours of battery life, and its small size makes it inexpensive to manufacture.
Aided by the new product, Intel’s gross profit on chips for notebook systems is about 75 percent of sales, 10 points higher than its margins on desktop PCs, estimates UBS Securities analyst Thomas Thornhill.
Intel is whipping up demand with a huge advertising blitz for a combination of chips called Centrino, which includes the Pentium M plus wireless networking and other accessory chips. The line’s strong sales are expected to help Intel’s gross margin — profits before operating expenses — to swell to 60 percent in the current quarter, up from 51 percent in the second quarter and near all-time peaks.
One sign of the new approach: Intel this year boosted the speed of its desktop microprocessors just once, compared with six times last year. Yet Intel has gained market share against chief rival Advanced Micro Devices. More Centrino-style product bundles are on the way, which Intel hopes will make it even harder for rivals to keep pace.
‘‘We are not playing the same game we played for 15 years,’’ says Paul Otellini, Intel’s president and chief operating officer, and the favorite to succeed Chief Executive Officer Craig Barrett.
The results are all the more striking because of the doubts swirling around Intel as recently as a year ago. Barrett had sank more than $10 billion into acquisitions in an attempt to lessen Intel’s reliance on the PC business, with little to show for it.
‘‘I don’t know of anything that we purchased that was worth what we paid for it,’’ says Andy Bryant, Intel’s chief financial officer. Its newer Itanium products, aimed at the largest computers, have won technical acclaim but few customers. And its cellphone and communications-chip units continue to lose money.
Meanwhile, Intel’s rivals are pushing hard. AMD in September introduced new desktop chips that are more powerful than Intel’s by some measures, a blow to Intel’s image as technology leader. Sun Microsystems is adopting a version of those chips in its server computers, as part of a technology partnership that could help AMD gain a stronger foothold in the corporatecomputing market.
In rethinking how it made and sold chips, Intel had to change the rules, because the old rules no longer worked. The focus on speed dated from an era when faster chips were needed regularly just to keep up with ever-more-complex software. By the late 1990s, however, most PC users no longer needed the hottest new chip for basic office tasks and surfing the Web. That meant they could hang on to their desktop machines for years, keeping down demand.