Data centers

12:35 PM
Connect Directly
RSS
E-Mail
50%
50%

Intel's Next CEO Must Reverse Years of Mobile Neglect

Intel built an unbeatable infrastructure for PC and server chips—which gives it zero advantage in meeting fundamentally different requirements for mobile chip design.

Whoever wins the soon-to-be-vacant CEO chair at Intel will presumably inherit a strong seat. Financially, Intel had a blowout year in 2011, in which both the top and bottom lines hit record highs. This year, its Data Center Group (server CPUs) is showing strong 5% year-over-year growth, and Intel dominates the server business, as our InformationWeek State of Servers report makes clear.

Sounds great, but Intel still generates more than three times as much revenue from the PC side of its business. That's a problem. The PC market, if not in outright decline, is undoubtedly in a state of long-term lethargy. And it's starting to show up in Intel's books. Its revenue and income have been flat this year, with a rough third quarter that can be easily blamed on procrastinating PC buyers waiting for Windows 8.

But even if Windows 8 is a hit, the future of client computing is mobile. By more than one estimate, global tablet sales will exceed sales of PCs by 2015, and, so far, Intel has been completely irrelevant to the smartphone and tablet markets.

Mobile devices present two fundamental challenges to Intel beyond eating into its fat PC margins.

First, mobile system requirements are dramatically different from servers. Whereas Intel could leverage the same chip architecture and design features for both its PC and server products, mobile devices are driving a wedge between client and server platforms. Even if Intel is ultimately successful in transforming its Atom line into a viable ARM alternative, its practice of copying and pasting many of the same subsystems from its Core products into server Xeon chips will come to an end.

Second, Intel was built on the notion of cranking out tens of millions of the same commodity parts. But despite largely using the same ARM CPU architecture, mobile chips aren't commodities, but rather a customized systems on a chip (SoC) built around a common, freely licensed core: the Dim Sum to Intel's mass produced Value Meal. Worse yet, some companies, most notably Apple, have built the CPU design talent to actually customize the stock ARM core, meaning the iPhone's A6 is tailor-made for that device.

As I wrote in September, Intel's mobile angst was on full display at this year's IDF, where the emphasis was more on power efficiency, mobile systems design and touchscreen UIs than it was on raw CPU performance, instruction set enhancements and faster system I/O. Yet if 2012 marked the year when Intel finally had its mobile epiphany, it might be too late to achieve salvation.

Don't think Wall Street hasn't noticed, as evidenced by this comment from a Piper Jaffray analyst: "While Intel's manufacturing capability is currently second to none, Samsung is catching up. Increasingly, new designs favor less expensive process technology, not high performance transistors used in PC CPUs. In the mobile era, customers such as Google, Facebook and Apple are designing their own mobile processors and server processors."

The problem for outgoing CEO Paul Otellini's legacy is that none of these market changes should have taken Intel by surprise: The iPhone is more than 5 years old and on its fifth generation, the first Android phone shipped over four years ago, and of course the iPad is coming up on its third birthday. All of this occurred on Mr. Otellini's watch, and while he's done a commendable job keeping Intel's tick-tock CPU strategy in perfect time, steadily improving both semiconductor process density and CPU architectural efficiency in alternating product cycles, he's arguably been oblivious as a classic disruptive technology came up and bit him in the rear.

Like Microsoft, Intel is facing multiple challenges to its lucrative PC franchise, and it may just be time for some fresh thinking at the top. In a world where smartphones and tablets are meeting a greater share of end users' IT needs, it's probably a good time for Intel to get new leadership; and between Intel's deep bench and the recent high-level departures at Apple and Microsoft, there's plenty of available talent.

Comment  | 
Print  | 
More Insights
Cartoon
Hot Topics
12
Confessions Of A VMworld Virgin
Susan Fogarty, Editor in Chief,  8/22/2014
10
Rethinking Data Center Design
Nicholas Ilyadis, Vice President and Chief Technical Officer, Infrastructure & Networking Group, Broadcom Corporation,  8/20/2014
6
IT Certification's Top 10 Benefits
Global Knowledge, Global Knowledge,  8/20/2014
White Papers
Register for Network Computing Newsletters
Current Issue
2014 Private Cloud Survey
2014 Private Cloud Survey
Respondents are on a roll: 53% brought their private clouds from concept to production in less than one year, and 60% ­extend their clouds across multiple datacenters. But expertise is scarce, with 51% saying acquiring skilled employees is a roadblock.
Video
Slideshows
Twitter Feed