Sticking with tradition, Intel has chosen an insider to be its next CEO—current chief operating officer Brian Krzanich. This might seem an unwise choice for a company whose last insider CEO led it over something of a cliff; sales shrank for the past two quarters and profits, while still a healthy 57%, continue to decline.
But here’s the thing: A close examination of Intel’s technology roadmap reveals that the company has a plan to recapture the crown as the world’s most prolific, profitable and technologically adept manufacturer of microchips. And it consists, primarily, of doing what the company has always done best—keeping its head down, focusing on engineering excellence, and investing heavily to make sure it is at least two years ahead of the competition in terms of the speed of its processors.
Speed is indeed key. Intel has always optimized its chips to be the fastest, and that’s why the company was tripped up by the mass shift of consumer demand from PCs to mobile devices, where speed is actually a problem, because it means more power consumption, too much heat, and unacceptably short battery life. A different company, ARM, which designs chips optimized for low power consumption, captured the lead and has a near monopoly on the main processors running all our mobile devices.
But there’s no reason Intel’s brain trust can’t start focusing on the metric that matters now—not raw performance, but performance per watt of power, pumped into a chip.
It’s an inelegant construction, but the decades-long monopoly of Windows software on Intel chips—known as “WinTel”—will, out of necessity soon be supplanted by a more nimble and promiscuous suite of collaborations between Intel and makers of mobile operating systems, i.e. Google. Call it GoogTel, Inoogle, whatever you like.
Intel has tried to get into mobile for a while now. Early versions of its power-sipping Atom chip made it into Netbooks, a disaster best forgotten. But now the company appears to be hitting its stride. What’s coming next are, according to an analysis by GigaOm, notebooks powered by Intel processors that run Google’s Chrome OS.
Depending on a person’s needs, Chrome OS is a legitimate competitor to Windows. (This is because most people’s computing needs, even on a notebook, are fairly modest—and can be answered exclusively by the cloud.) And Chrome OS is not only lightweight and fast, it’s free, which takes the Microsoft premium off any notebooks that carry it—one reason why Samsung’s $250 Chromebook (which features an ARM-based processor) is remarkably functional, despite its low price.
According to mobile editor Kevin Tofel at GigaOm, a Chromebook running one of Intel’s forthcoming “Haswell” processors, which has been optimized for low power consumption, could have a battery life of up to 10 hours. That kind of power combined with that kind of low power consumption points to a fundamental advantage Intel has over its competitors, that the company will soon exploit in its mobile-only Atom chips: Smaller features on Intel’s microchips, which automatically lead to smaller chips and lower power consumption.
Intel is at least a year to 18 months ahead of competitors on producing mobile chips with features at the “22 nanometer” resolution, which means individual elements on the chip are just 22 nanometers wide. And by 2014, the company claims it will already be manufacturing chips at the next-smallest “process” step, which is 14 nanometers, for even better performance. If Intel delivers, this could mean processors that are significantly faster than the ones from competitors such as Qualcomm, Samsung and Apple, all of which use either ARM chip designs or the ARM instruction set.
Intel’s competitive advantage is its expertise in manufacturing. Unlike all other high-end processor manufacturers save Samsung, Intel owns its own factories—called fabs—for making microchips. These are multi-billion-dollar operations that companies like Qualcomm simply can’t afford to have on their balance sheets. (If Apple, which currently relies on Samsung and Taiwan Semiconductor to produce its chips, ever built a fab, that would be a game-changer.)
This is where Krzanich comes in. If Intel’s future advantage is the same as its old advantage—expertise in the incredibly difficult, capital-intensive game of making new semiconductor manufacturing facilities, the company needs an insider with decades of institutional knowledge to preserve that advantage. An outsider might have accelerated Intel’s shift into new businesses—more mobile chips, or becoming a contract manufacturer for other chip designers, say—but the company’s latest move suggests it’s already on the right path. All it needs is a steady hand at the till, someone who can preserve Intel’s current edge.
A newly re-focused Intel, still a technology leader but finally producing the power-sipping chips the market demands, is a company ripe for a comeback. The company’s share of the mobile chip market is still in the low single digits. If Haswell, future versions of Atom, and their successors offer notebook, tablet and phone manufacturers a way to differentiate themselves with more capabilities than current ARM-based chips, you can bet they’ll make the shift to Intel. And that could help Intel start to make up for the shortfall in sales of PCs and servers.