A version of this article appears in the October 28, 2013 issue of Fortune.
On a Saturday morning in late April, members of Intel’s board of directors settled into a meeting room at the Ritz-Carlton in Half Moon Bay, Calif., for a top-secret gathering. Flanked by security guards stationed at the ends of the hallway, they readied themselves for the onslaught of PowerPoint presentations they’d soon be viewing, querying, and analyzing.
The board’s search for a new CEO had already lasted several months and included multiple waves of interviews and deliberations. This was the last stretch, a lightning round of sorts. Led by chairman and former Intel chief administrative officer Andy Bryant, the 10 directors had narrowed down the hunt to three contenders: two longtime company insiders and one external candidate. Today, each of the finalists would get another shot to make his or her case. The winner’s prize? Steering the chipmaking behemoth through its toughest transition yet—the post-PC era.
The Intel (INTC) of today is a $53.3-billion-in-sales tech giant whose immense factories churn out a collective 1.5 million chips per day. Eight out of 10 PCs in the world run on the company’s powerful and high-profit-margin processors. The only problem is, demand for PCs is waning fast, despite Intel’s repeated efforts to reinvent the category. Last year global PC unit shipments dropped 4%, according to IDC. By the end of 2013 that number is expected to fall nearly 10% more.
For more on Intel, watch this Fortune video:
It is as simple as this: The world has gone mobile. PCs are passé. And the popularity of smartphones and tablets has spawned whole industries of mobile applications, devices, operating systems, and chip architectures. Global tablet shipments are expected to eclipse PCs by 2015.
Unfortunately for Intel, it has amassed less than 1% market share in tablets and phones, according to estimates from Sanford C. Bernstein & Co. “They are not the incumbent in mobile,” says Jim McGregor, founder of Tirias Research and an analyst who has tracked Intel for decades. “They are a company that’s trying to break into the largest ecosystem in technology today.”
Whoever the board picked in Half Moon Bay would inherit a company on the cusp of a destabilizing and culture-shifting change, one that threatened to alter the foundation of Intel’s entire business model. And in many ways, Intel’s CEO hunt was an apt metaphor for its macro struggles: The company had been caught off-guard by the speed of the transition—and now its options were, well, limited.
In late 2012, then-chief executive Paul Otellini announced, to the board’s surprise, that he would step down in the spring, two years before his expected retirement. Like almost everything else at the famously disciplined semiconductor company, succession planning had long been a carefully orchestrated affair, but this time around, the pool of ready-to-go candidates was narrower than usual, says a person close to the board. There was also serious pressure to consider external CEO candidates—an unprecedented move by Intel. Promoting from within was an idea steeped in Intel’s culture. But some directors worried that Intel’s tardiness to the smartphone and tablet markets had been due at least in part to its insularity.
Then came a more unusual wrinkle. The two internal finalists—former COO Brian Krzanich (pronounced kr-ZAN-itch) and former software head Renée James—declared they would make their board presentation together. Several weeks prior, they’d come to the conclusion that collaborating, rather than competing, would increase the chances of the board’s picking one of them and not the candidate they referred to as “Outside Guy.” (Intel won’t confirm it, but VMware CEO Pat Gelsinger and Sanjay Jha, the former CEO of Motorola Mobility, are thought to have been two external contenders for the CEO job.)
“Our image of where the company should go was almost identical,” says the 53-year-old Krzanich, in a joint interview with James at Intel’s Santa Clara, Calif., headquarters. “It was us against the outside.” James, 49, echoes the sentiment: “There was no possibility in our point of view that anyone could come in here and know what to do,” she says. “We agreed that they pick both of us or none of us.”
Krzanich, known within Intel as B.K., had spent three decades at Intel, mostly on the manufacturing side. James had joined Intel in 1988 but had worked her way through the ranks via another route—leading software subsidiaries like McAfee and Wind River. The two executives had spent weeks developing and prepping for their final presentation to the board. Their game plan was straightforward: They’d flatten Intel’s organization by having the general managers of the PC group and other product divisions report directly to the CEO. They’d ramp up production of Atom, the company’s line of lower-power processors for tablets and smartphones, by redirecting resources away from the powerful Intel Core processor family. And they’d focus on integrating more mobile-friendly components, like graphics and communications chips, into their product line for phones and tablets.
When the presentation was over, the board grilled each candidate separately. That evening the 10 directors began deliberating in a private dining room at the Ritz, broke for the night, then continued the heated discussion the following morning. Around noon on Sunday, Krzanich received a text message from board chairman Bryant asking him to return to Half Moon Bay. Krzanich, who was at his daughter’s championship basketball game, hadn’t expected to receive an answer for another day or two. In a T-shirt, shorts, and sneakers, he drove back to the oceanside hotel, where he learned he would become Intel’s sixth CEO. “You’re numb,” he recalls. “You want it, and you’re scared of it. You don’t expect it. And then you’re just sitting there thinking about what you should do first.”
What he did first was to ask the board to make James president. (In 2013, she debuted on Fortune‘s Most Powerful Women list at No. 27.) In truth, the move was not only a thank-you to James, but also a nod to Intel tradition—before Otellini’s time, Intel’s upper management had always been a two-person job. James would report to Krzanich, but the entire company would report to both of them. Krzanich was awarded an annual salary of $1 million. Factoring in a cash bonus and stock options worth $6.5 million, he stands to earn upwards of $10 million in 2013.
But certainly a bigger nod to Intel tradition was the choice of Krzanich himself. Krzanich was a logical pick for CEO. (Both Otellini and his predecessor, Craig Barrett, had also served as COO before getting the top job.) He had begun his career as an engineer at the company’s facilities in Albuquerque, eventually working his way up to plant manager in Intel’s Massachusetts-based fab (i.e., semiconductor factory). Later he had been named senior VP of all manufacturing operations, before being bumped up to COO in January 2012.
Along the way, he has earned many fans both within and outside Intel’s walls. “He’s very cool, thoughtful, calm, and decisive,” says high-profile venture capitalist and former Intel sales exec John Doerr, who once served on the board of MiaSolé, a solar company, alongside Krzanich. But importantly, Krzanich is also a PC guy—something that the new boss acknowledges is woven into the company’s core, and presumably its longtime execs too. “This company spent 20 years in a battle with AMD [its top competitor in the PC business] where a half a megahertz in performance made the difference [between positive or negative reviews and sales],” Krzanich explains. “Everything in our structure, our DNA, is centered on that.”
Some on the board are no doubt hoping that the partnership with James will mutate some of that genetic predisposition. Intel’s new president, after all, is something of a novelty at the Santa Clara institution. That’s not just because she’s the highest-ranking woman in the company’s history. She also came out of the software side of the business, unusual for Intel’s higher-ups—though it’s a background that may be essential if Intel’s products are to play nice with mobile-phone components, especially on newer operating systems like Google’s Android.
As a woman in a male-dominated culture once described as “a bunch of aggressive introverts,” James has learned to work her way up. “It has been hard,” she admits. “But not hard because of the things you would think were hard. It wasn’t that I didn’t get opportunities or great access to mentors. It was the subtle, unsaid gender bias that if you’re tough and you’re a woman, you’re a bitch. If you’re tough and you’re a guy, you’re a hero.”
For more on Intel, watch this Fortune video: [fortune-brightcove videoid=5032321041001]
At once bubbly and brash, she comes across more as Krzanich’s peer than as a direct report. Early in her career at Intel she was handpicked by Andy Grove, Intel’s former CEO and legendary leader, to be his technical assistant—a fast track up Intel’s corporate ladder—where she served for more than four years. “She is bullheaded, she does not stop, she is impatient and very smart—that goes without saying,” Grove, now 77, says of James. “I envy her for her energy.”
Still, not everybody thinks such energy will be enough to breathe new life into Intel. Many investors, indeed, had hoped for more of a management shakeup. “While Krzanich will protect the company’s crown jewels,” Piper Jaffray analyst Gus Richard wrote in a blunt note to clients the same day Intel’s CEO pick was revealed, “he does not have the background that would suggest he has the market acumen to lead the company during a period of tectonic change.”
Krzanich and James now have one job, it seems: proving that assessment wrong.
When analysts—or make that any student of business—talk about leading in a time of tectonic change, they’re often alluding to an Intel executive: Andy Grove. It was Grove who liked to say that Intel was one wrong answer away from disaster. His book, Only the Paranoid Survive, published in 1996, became a managerial guidebook for companies grappling with massive, sometimes debilitating, market changes. In the early 1980s, when Grove was president of Intel, the company’s core business, memory chips, tanked because of competition from Japanese manufacturers. Grove suggested the unthinkable—instead of competing to the bitter end, Intel should scrap the memory-chip business altogether and bet the company on a new product, microprocessors. The shift was brutal—in 1986, Intel fired about 7,000 people and lost more than $170 million on $1.3 billion in sales—but Grove eventually persuaded computer maker IBM (IBM) to stock its emerging PC product line with his company’s microprocessors. Intel’s business surged to unprecedented proportions.
To borrow from Grove’s lexicon, today’s Intel is once again facing a “strategic inflection point.” This time around, the company is struggling to figure out how to swing the transition to its advantage. “Mobile is a monster trend,” says Doerr. “It’s the biggest thing that’s happened in the tech world in my lifetime.”
It’s not that Intel hasn’t tried to crack the mobile market. But recently it’s had many more flubs than successes. Under Otellini the company repeatedly made promises it couldn’t keep. In early 2009, Intel proclaimed that the Atom-based Moorestown system-on-a-chip would be featured in smartphones the following year. But an LG Electronics-made mobile device was later scrapped, and the chip made it only into netbooks and robotic products. In other unfortunate moves, in 2010, Otellini decided to partner with Nokia to develop MeeGo, a Linux-based operating system for phones. A year later Nokia ditched the project and hitched its future to Microsoft’s Windows Phone instead. (The Finnish phonemaker has agreed to sell its device business to Microsoft.) Efforts to invest in and integrate more mobile-friendly features were repeatedly “ZBB’d”—short for zero-based budgeting, a term Intel insiders use to describe defunding projects. “If you had the keys to a kingdom that generated $55 billion in revenue, wouldn’t you worry about being the one who caused the cratering of this jewel?” Krzanich says of Otellini’s reign.
For a different take on zero-based budgeting, see “Squeezing Heinz.”
How could such a cutting-edge juggernaut miss out on such a large, growing market? There are several reasons, but as Krzanich implies, it mainly comes down to this: Intel is a victim of its own success. As it busied itself with milking profits from the PC industry and keeping up with Moore’s law by investing in the most advanced chipmaking technology, the world around it changed. The mobile market grew more rapidly than Intel anticipated, and with it came demand for low-power, not just high-performance, chips. And while Intel’s processors were originally designed for machines that plugged into a wall, newer, competing chip designs were created for smaller, sleeker mobile devices. They could enable longer battery life and incorporated other functions important for mobile use: GPS, broadband standards like 3G and LTE, and short-range technologies like Bluetooth and near-field communications.
Intel spent years fighting rival AMD for the upper hand in PC and server chips, but the mobile ecosystem has many, many more levels of competition. British company ARM Holdings licenses the underlying chip blueprints. So-called fab-less semiconductor makers such as mobile powerhouse Qualcomm and graphics specialist Nvidia employ Asian chip factories to manufacture their goods. Because Intel operates its own fabs and relies on a different chip architecture, called x86, it has been virtually locked out. The most popular phones and tablets—like Apple’s iPhone and iPad, for example—all run on ARM-based chips, not Intel.
Such concerns have weighed heavily on the stock. Over the past 10 years Intel’s share price has dropped nearly 23%. In the same period Qualcomm’s stock has surged by 200%, and Nvidia’s by 177%. ARM’s share price, meanwhile, has ballooned over 800%. (Intel’s market cap of $114 billion, to be fair, dwarfs ARM’s $22 billion.)
“Now we’re in a world where the software runs on ARM,” says Simon Segars, ARM’s CEO. “The barrier to entry is much lower.”
Intel’s new management concedes that the company was slow to recognize the shift to mobile. “I think there was a mentality that we would come back—that we’d seen downturns before and we can innovate in the PC and win these people back,” says Krzanich, back at the Bay Area headquarters with James. “Our dilemma now is time.”
Hours after their coming-out at Intel’s annual stockholders’ meeting on May 16, Krzanich and James invited the company’s top dozen executives to an offsite meeting at the Rosewood Sand Hill hotel in Menlo Park. During the next day and a half, the management team identified eight “strategic imperatives.” Everyone was given an assignment and asked to come back the following month to present his vision for how the proposed changes should be carried out and how results should be measured.
The following week Krzanich sent out an internal memo outlining significant structural changes at Intel. Just as Krzanich and James had laid out in their presentation to the board, most of the main product groups were to be placed directly under the CEO’s supervision. Longtime Intel architecture chief Dadi Perlmutter was removed from that role to enable the shift. The vision? Speed up lines of communication—and drive accountability—in an effort to move more quickly and decisively in smartphones and tablets. “Paul [Otellini] did fabulous things for this company,” says James, “but we made decisions from the point of view of what had made us successful in the past. In mobile, different things are important.”
Intel’s Atom-based Bay Trail platform has made some headway in tablets and finally garnered the company some good reviews in more mobile, lightweight products. Electronics manufacturers like Toshiba, Asus, and Lenovo are all developing Bay Trail-based tablets, the first of which will be available for the holiday season. (Prices start around $200, though Intel says a sub-$100 Atom-powered tablet isn’t far off.)
But even as Intel scrambles to catch up in mobile, success on the tablet front in particular is likely to help cannibalize the higher profit margins it has grown accustomed to in the PC market. Plus, it risks missing out on the next big trend—think smart watches and Internet-connected glasses.
That’s why, on Sept. 10, at Intel’s annual developer conference in San Francisco, Krzanich and James took the stage to announce an upcoming family of processors called Quark, specially made for wearable devices and industrial “Internet of things” applications. The new Quark chips, Krzanich told a crowd of several thousand people, will be five times smaller and 10 times more power-efficient than Intel’s Atom chips. Later, both Krzanich and James took questions from the audience, an unscripted (and unheard-of) move for Intel. At first no one came forward. But gradually developers lined up to ask about Intel’s new strategy, product road map, and the possibilities for Quark. “It demonstrated Intel has the ability to build ultralow-power products,” says Doug Freedman, an analyst with RBC Capital Markets. “And it showed their willingness to grab products from the labs and productize them, whether fully baked or not.”
Even if Quark succeeds, the wearables market will probably take a few more years to scale, and it’s not clear whether Intel will ever gain a leadership position in the new, more mobile computing form factors that now dominate our attention spans and wallets. In the meantime the company’s biggest problem is keeping its factories running at capacity. Intel’s most important asset is also its biggest Achilles’ heel—fabs cost billions of dollars to build, operate, and maintain. Their financial viability is predicated on Intel’s ability to keep driving demand for the teeny-tiny chips they produce en masse. According to McGregor, the Tirias Research analyst, Intel’s fab utilization rates have been running well below “normal” levels. (Intel declined to comment.)
To keep its fabs humming, Krzanich and James are exploring another revenue driver: customized chips for other companies. It was actually Otellini who started Intel down the path of opening its fabs to outsiders. But Intel’s new management duo has already hinted that they’ll be more open-minded about going into the so-called foundry business—even if it means manufacturing competitors’ chips. And while that’s unlikely to be an easy transition either, it might be an inevitable one. “Moore’s law—smaller, faster, and cheaper—has been the cornerstone of tech economics for 40 years,” wrote Credit Suisse’s John Pitzer in a recent report. “It is our view that technical, financial, and scale issues are driving toward a ‘last-man-standing’ advantage for Intel relative to Moore’s law, and that Intel will leverage that advantage not only in its core business but increasingly as a specialized foundry.”
Of course, sharing its manufacturing prowess with the competition—even if it helps Intel’s bottom line—is the ultimate admission that its x86 architecture may never reign supreme again, at least not when it comes to today’s personal computers of choice. It’s also an acceptance that Intel’s business structure—and the fat margins it enjoyed in the PC era—may need to change, in part because the foundry business simply isn’t as profitable. Whether that’s a leap Krzanich and James, however open-minded, can make remains to be seen. “What we realized from the beginning was that you had to let go,” Krzanich says. “Instead of trying to bring everybody back [to the PC], you’ve got to embrace it. It’s a risky situation, but when you bring the assets of Intel with you, it isn’t so scary to embrace it.”
It would be foolish to rule Intel out. The company has faced near-death situations in the past and has come out stronger. But the more successful (and complacent) a company becomes, the harder it is to admit defeat in markets that it should—and could—have dominated and to make the switch quickly to plan B. For Krzanich and James, that’s likely to mean a few more years of painful changes. Luckily for the new CEO and president, they’re already well versed in a culture that has revived itself before.