Intel's New Atom: Moorestown and Beyond
May 18, 2010 5:00 AM PT
Intel's next-gen Atom processor-based platform (formerly "Moorestown") caused quite a stir in the news and among mobile computing aficionados. On the technical side, Intel seems to have delivered the goods. The platform includes Intel's Atom processor Z6xx Series Family (formerly "Lincroft"), the Platform Controller Hub MP20 (formerly "Langwell") and a dedicated Mixed Signal IC (MSIC) (formerly "Briertown").
It adds 3-D graphics, video encode and decode, and memory and display controllers into the single system-on-chip (SoC) design. Also included are the MP20 Platform Controller Hub and a dedicated MSIC, integrating power delivery and battery charging, and consolidating a range of analog and digital components.
What does this mean in plain English? That the new Atom platform is simply the best performing solution for handheld computing devices Intel has ever developed.
How much better? Try >50x reduction in idle power, >20x reduction in audio power and 2-3x power reductions in browsing and video compared to the previous-generation Atom (aka "Menlow") processor. These efficiencies translate into >10 days of standby, up to two days of audio playback and four to five hours of browsing and video life with common 1500mAh batteries.
Compared to Menlow, Intel hit the ball out of the park with the new Atom, most analysts seemed to agree, but many were less impressed with the platform's commercial prospects. That was not particularly surprising given the turmoil in the mobile marketplace that began with Microsoft's February announcement of its Windows Phone 7 Series OS, followed by the weeks of hype leading up to Apple's April iPad launch and its acquisition of Intrinsity, and HP's purchase of Palm and the reported cancellation of its Windows 7-based Slate tablet.
Given those events, the conventional wisdom on Intel's new Atom platform appears to be that 1) Apple's dominant mindshare in all things mobile makes it the company to beat; 2) close relationships with Apple and smartphone market leader RIM make ARM-based microprocessors virtually unstoppable; and 3) despite impressive improvements, Intel's new Atom is simply too late to the mobile game to be a viable player.
So is the conventional wisdom particularly wise? I have my doubts. Certainly Apple, as well as leading smartphone vendor RIM, deserves applause for some terrific product and service offerings. Apple, in particular, has seized the public imagination regarding the smartphone and tablet experience to a remarkable extent. Plus, the company's App Store has defined the critical role a deep application portfolio plays in platform success.
However, recent APD Group research suggests first-quarter sales of Android-based phones blew past iPhone sales. If Android's success continues, it could indicate that Apple's single-vendor development and manufacturing model faces significant barriers in the global market for handheld computing.
In addition, though the iPad has enjoyed extraordinary sales to date, it is the lone occupant of an essentially nascent market that will be crowded with enticing new products by the holidays. Some of these will almost assuredly be based on the new Intel Atom platform.
Worth the Investment
What about ARM? Can Intel really hope to prevail against so dominant a competitor? That's certainly problematic, especially given ARM's deep client roster. However, the sizable semiconductor investments of one of its most important customers -- Apple -- could mean future problems for ARM. More importantly, a core (no pun intended) Intel strength is its ability to play the long game, as was clear in the company's response to AMD's Opteron processor.
How so? AMD's singular pursuit of 64-bit x86 computing solutions allowed the company to develop a sizable market and mindshare lead. Though it initially dismissed Opteron, Intel eventually dove into the market with a vengeance and eventually left AMD in its wake. That point is worth remembering in still-evolving markets like mobile computing.
The new Atom platform demonstrates, yet again, Intel's ability to effectively adapt to changing circumstances by leveraging its considerable intellectual and human capital.
This is not to say that the new Atom platform will be a slam-dunk success. Mobile computing is far less homogenous than the PC, laptop and server markets, with a technologically diverse vendor ecosystem and highly (some might say wildly) differentiated global customer groups. In other words, it's a world that will be harder to engage with and work within than the more predictable PC/server networks Intel is used to.
Yet that same world is also populated with dynamic, astute vendors and customers willing to embrace new innovations. Bottom line: Known and unknown challenges aside, I believe the potential opportunities for the new Atom platform are enormous, and worth every bit of Intel's time, effort and investment.
E-Commerce Times columnist Charles King is principal analyst for Pund-IT, an IT industry consultancy that emphasizes understanding technology and product evolution, and interpreting the effects these changes will have on business customers and the greater IT marketplace.