By Peter Burrows About midway through his keynote speech at Apple's (AAPL) Worldwide Developers Conference in San Francisco on June 6, Steve Jobs told the audience what it already knew. "It turns out that the rumors were true," he said, proceeding to lay out the details of Apple's plan to drop the PowerPC chip from IBM (IBM) and Motorola spin-off Freescale semiconductor and build future models around Intel's (INTC) omnipresent microprocessors.The Mac OS has been leading a "secret double life" for the last five years. That's when Apple engineers first began working to get the Macintosh operating software to run on Intel silicon.
That bit of history was one element in a slew of details aimed at limiting the nervousness of Apple's partners, customers, and investors. While computer outfits must upgrade to new chips every decade or so to remain competitive, it's a hugely treacherous endeavor. Not only can an execution problem bring sales to a standstill but even news of the transition can stifle sales of existing models.
Apple, which switched from Motorola's (MOT) 68000 chip to the PowerPC in the early 1990s, presented the outline of a cohesive vision for accomplishing this latest transition. "I think today is a historic event for Apple, but not a disruptive one for its developers or customers," concludes JupiterResearch analyst Michael Gartenberg.
DEMANDING DESIGNS. Apple's migration to the Intel fold could extend its current Golden Age by boosting Mac sales and by enabling it to reprise its success with the iPod and create new consumer devices. In part, that's because having the "Intel Inside" sticker on its products -- figuratively if not literally, in case that logo doesn't meet with Apple's discriminating sense of style -- could help Jobs & Co. win customers who have been worried about Apple's traditional go-it-alone approach.
The substance of this deal, however, is about Apple's desire to tap Intel's chip expertise, not its marketing clout. While long-time partner IBM focuses on making the speediest chips possible -- it specializes in producing some of the world's most powerful computers -- Intel focuses on chips that combine maximum speed with the lowest possible heat. That's a perfect fit for the kinds of thin, portable, minimalist Macs that Jobs probably has in mind. "As we looked down the road, we simply didn't see how we were going to be able to make the products we want to make" using PowerPC chips, Jobs said in an interview.
He claims Apple already is far along in making the Mac OS run well on Intel chips. It turns out that Apple kept working on this problem after bringing Jobs back into the fold when it bought his former company, NeXT Software, in 1996. That outfit's operating system, NextStep, ran on Intel chips.
CHECK THE BOX. After using that code to create Apple's new Mac OS X operating system in 2000, Jobs insisted that all future versions be able to run on either PowerPC or Intel. Analysts and others at the conference saw convincing evidence during Jobs's keynote presentation, which he ran on an Intel-based Mac. After his talk, developers were surprised that the machine had been able to simultaneously run many different programs, including memory hogs like Adobe (ADBE) Photoshop.
That solves only part of the problem, however. Apple has to get software developers to create programs for both PowerPC and Intel-based Macs, since Jobs says the transition to Intel won't be complete until the end of 2007. Toward that goal, Apple demonstrated software tools that allow developers to carry out this "recompile" by simply checking off "Intel" in an on-screen dialog box.
While it may not prove quite that easy in practice, Jobs conferred credibility to the concept by having Wolfram Research software designer Theo Gray recount how his company was able to recompile its complex Mathematica program in less than two hours. Jobs also encouraged developers to use Apple's "universal binary" technology to create CDs of their programs capable of running on either PowerPC or Intel-based systems. That way, customers would just install the CDs on any Mac, and the right version of the code would automatically load.
LOST SHARE. Jobs also addressed the issue of enabling existing Mac customers to upgrade their programs so they'll continue to work with new models. Apple announced a piece of code, called Rosetta, that will let even existing programs written for PowerPC-based Macs run on Intel-based machines, albeit with some negative impact on processing speed.
Major risks remain. For starters, Apple intends to introduce its first computers based on Intel's processors within a year, and convert remaining product lines by the end of 2007. That leaves a lot of time for rivals such as Dell (DELL) and Hewlett-Packard (HPQ) to generate "fear, uncertainty, and doubt" about Jobs & Co.'s ability to pull off its plan.
Indeed, Apple's market share fell from 10% in 1992 to 5% in 1994, the years of its transition to the PowerPC, says Nathan Brookwood, an analyst with consultancy Insight 64. While Apple had other serious problems during that period, uncertainty surrounding the processor change certainly didn't help.
PERIOD OF UNCERTAINTY. Apple is bound to lose some of its customers as this migration takes place. Some users won't want to worry about which software is compatible with which Mac models -- or whether those systems will work reliably. "I don't want to play this game," says Joon You, chief technology officer for contract researcher Target Health in New York, which just bought its latest Mac last week. "I have so many other things to worry about."
Instead, You plans to stop buying Macs. He also plans to stop evaluating Apple's servers and go with Sun Microsystems' (SUNW) products. While Sun also recently began a migration to Intel for some of its wares, it's already far down that path. Apple, on the other hand, is just entering a period of uncertainty. "I think Apple made a great mistake," You says.
An even greater risk in the short term is the possibility that significant numbers of Mac buyers will hold off on making their purchases, waiting for the opportunity to get their hands on an Intel-based Mac instead. The problem could be especially dangerous for laptop sales. Already, Apple PowerBooks run more slowly than Windows-compatible models. A top-of-the-line PowerBook runs at 1.67 gigahertz, vs. 2.13 gHz for a high-end Dell.
QUEST FOR VIBRANCY. Then there's the question of how Mac zealots might react. "Since its inception, Apple has profited hugely from a customer base whose dedication often bordered on religious devotion," writes Charles King, an analyst with Pund-IT Research, in a recent note. "Many members of that group tend to deeply distrust the concept of computing promoted by Intel and its long-time partner Microsoft (MSFT)."
Still, Jobs got off to a good start in addressing these many fears -- and he and Intel CEO Paul Otellini haven't even begun to lay out the ways they can improve the Mac. Asked whether Apple can better create innovative products than cost-conscious PC makers such as Dell, Otellini responded carefully: "Having a customer that can match our intensity [for innovation] is really exciting for us. The industry needs really interesting new products to keep it vibrant."
Over the next two years, Otellini will have plenty of opportunity to justify his excitement. With Olga Kharif in Portland, Ore., Steve Hamm in New York, and Cliff Edwards in San Mateo, Calif.
Burrows is Computer editor in BusinessWeek's Silicon Valley bureau