Intel Talks Tablets, Reinvents PC, Avoids iPad
At its financial analyst meeting earlier this week, Intel unveiled a series of strategic changes to its roadmap and gave investors a peek at the company's general strategy for the next few years. We've already touched on the company's decision to ramp more powerful Atom processors and its new, 15W mainstream notebook TDP target, but the company is playing a deeper game.
If Medfield lives up to expectations, it'll be the first Intel SoC truly capable of challenging ARM.
Later this year, Intel will introduce Medfield, its 32nm Atom SoC. The current solution, Moorestown, did reduce Atom's platform-level power consumption, but not enough to make the chip attractive to smartphone designers. Medfield is meant to change that; the company demonstrated Medfield devices at the meeting and showed a slide on how the 32nm SoC's power consumption compares to the current competitive range.
As for the PC market as a whole, Intel doesn't believe tablets pose a significant threat to the "traditional" mobile/desktop market. The catch, however, is that Intel is moving to redefine what constitutes a PC even as it defends the long-term stability of the supposedly stolid segment. Intel's next-generation conceptual system owes little to current systems and greatly resembles a Macbook Air. The company's decision to focus mainstream production on 15W TDP processors as opposed to the current 35W standard reflects Intel's desire to 'save' the PC market by adopting the diminutive form factor and weight of a tablet while providing higher performance, a physical keyboard, and additional ports for video, USB peripherals, and an optical drive.
Intel's Atom roadmap through 2014. Note that Atom won't debut on a cutting-edge process until 2014, but the product's ramp curve accelerates dramatically in 2012-2013.
We suspect Intel's renewed focus and redefinition of the PC market is due in part to the dearth of x86 tablet solutions. Nokia and Intel are still working on Meego, but the former's prominent decision to standardize on Windows Phone 7 going forward dramatically reduced the resources and focus the Finnish company was willing to commit to the open-source operating system. Windows 7's unoptimized-for-fingers UI makes it a bad joke as far as tablets are concerned, while the Android x86 project isn't ready for prime time yet.
The iPad was conspicuous by its absence from the discussion, Intel never even referred to the device. Apple's design to build customized ARM cores has locked the CPU manufacturer out of what's arguably the only successful tablet currently on the market. If Apple's interested in switching ARM for x86, the company's done nothing to indicate it—the A5 processor that powers the iPad 2 will undoubtedly scale to higher performance and lower power consumption as companies like TSMC move to 28nm.
Then again, Jobs has repeatedly demonstrated that he's more interested in competitive hardware than he is in CPU branding. Having an in-house silicon design team is a feather in Apple's cap, but the company would likely switch to Intel's Atom if that processor demonstrated a better performance-per-watt ratio than what Apple felt it could achieve.
Intel's redefinition of the PC market will have a significant impact on the shape of things to come. The company isn't abandoning notebook users who prefer more powerful systems that fit the traditional 35W envelope, but its focus on low power consumption and extended battery life could impact the amount of R&D various companies devote to these two pursuits. This would be particularly welcome where batteries are concerned.
According to the website Battery University, "the energy density of modern batteries improves by about 10 percent per year." The reason we don't see that additional capacity making a tremendous difference over time is that device power consumption has typically increased just as fast (if not faster) than energy density does. Intel's new focus could put the brakes on that trend and lead to better runtimes for everyone. Atom has its place, but most users would prefer the power of a higher-end processor and an Atom-equipped netbook's battery life.
If Medfield lives up to expectations, it'll be the first Intel SoC truly capable of challenging ARM.
Intel's Atom roadmap through 2014. Note that Atom won't debut on a cutting-edge process until 2014, but the product's ramp curve accelerates dramatically in 2012-2013.
We suspect Intel's renewed focus and redefinition of the PC market is due in part to the dearth of x86 tablet solutions. Nokia and Intel are still working on Meego, but the former's prominent decision to standardize on Windows Phone 7 going forward dramatically reduced the resources and focus the Finnish company was willing to commit to the open-source operating system. Windows 7's unoptimized-for-fingers UI makes it a bad joke as far as tablets are concerned, while the Android x86 project isn't ready for prime time yet.
The iPad was conspicuous by its absence from the discussion, Intel never even referred to the device. Apple's design to build customized ARM cores has locked the CPU manufacturer out of what's arguably the only successful tablet currently on the market. If Apple's interested in switching ARM for x86, the company's done nothing to indicate it—the A5 processor that powers the iPad 2 will undoubtedly scale to higher performance and lower power consumption as companies like TSMC move to 28nm.
Then again, Jobs has repeatedly demonstrated that he's more interested in competitive hardware than he is in CPU branding. Having an in-house silicon design team is a feather in Apple's cap, but the company would likely switch to Intel's Atom if that processor demonstrated a better performance-per-watt ratio than what Apple felt it could achieve.
Intel's redefinition of the PC market will have a significant impact on the shape of things to come. The company isn't abandoning notebook users who prefer more powerful systems that fit the traditional 35W envelope, but its focus on low power consumption and extended battery life could impact the amount of R&D various companies devote to these two pursuits. This would be particularly welcome where batteries are concerned.
According to the website Battery University, "the energy density of modern batteries improves by about 10 percent per year." The reason we don't see that additional capacity making a tremendous difference over time is that device power consumption has typically increased just as fast (if not faster) than energy density does. Intel's new focus could put the brakes on that trend and lead to better runtimes for everyone. Atom has its place, but most users would prefer the power of a higher-end processor and an Atom-equipped netbook's battery life.
