Here's Why You Should Cut Intel Some Slack

Back at Mobile World Congress 2014, Intel launched its next-generation smartphone platform, codenamed Merrifield. At the heart of the platform are a dual-core Silvermont processor and Imagination Technologies PowerVR G6400 graphics, and it weighs in at about 70 millimeters squared on the die size (suggesting that its manufacturing cost is perhaps just $5). The modem that will initially come with the platform is Intel's first-generation XMM 7160, but customers could choose to pair it with the much-improved XMM 7260. Nearly two months following the launch, not a single design win has been announced. What gives?

Not being in the loop isn't fun
The development cycle for a mobile processor is about four years, suggesting that when Intel's marketing group was deciding what features to put into this processor, a "smartphone" was basically the 3.5-inch iPhone with a 480x320 display, a single-core ARM Holdings Cortex A8 at 600MHz, and an Imagination Technologies PowerVR SGX535.

Now, when you're Apple, the company that leads the industry, it's easy to know what kind of processing power your phone needs -- which is why Apple builds its own chips. Or, if you're Qualcomm -- the leading smartphone chip vendor today -- you're working closely with heavyweights like Samsung and LG to understand exactly what the requirements are for next-generation phones. But what if you're Intel, new to this industry and with no "friendly" customers to tell you what to design?


It must've seemed crazy fast at the time
Back in 2009, when Intel was likely drafting the initial specifications for the Merrifield platform, the idea of two Silvermont cores -- cores significantly faster than the best netbooks available at the time -- in a phone must've seemed like a pretty significant leap. After all, the ARM Cortex A8 at 600 MHz couldn't hold a candle to the 1.6GHz+ Intel Atom at the time.

Indeed, two Silvermont cores running at 2.13GHz max turbo (the speed of the top-bin Merrifield, known as the Atom Z3480) are pretty dramatically ahead of what the iPhone 3GS had to offer and probably seemed like a reasonable design point to target for leadership in the smartphone market for a 2013 rollout. Unfortunately, Intel underestimated just how quickly the ARM folks (as well as Qualcomm) would be able to scale up the performance levels of their processors.

Hey, notice how phones got really big all of a sudden?
With Intel likely aiming for an iPhone or an iPhone-class design, there was another thing the company probably didn't count on when it first sat down to develop the Merrifield platform -- bigger phones. Larger phones, like the Samsung Galaxy S and Galaxy Note products, offer larger batteries and allow the processor to dissipate more power, particularly for short bursts of time.

Intel's reasoning behind Merrifield was probably an aim to be able to sustain fairly high performance for long periods. A dual-core Silvermont (Silvermont dissipates less than 1 watt per core at maximum load) would be able to peg top frequency for quite some time without throttling back. Contrast this with, say, the quad-core Qualcomm Snapdragon 800, which -- as Anandtech showed in testing the Nexus 5 -- throttles back from its peak advertised 2.3GHz to about 1GHz under heavy load.

Technical elegance has given way to marketing
Look at some of the ridiculousness in the mobile market today, with "octa-core" processors, big.LITTLE, and other technical abominations that at best serve as marketing points. Most mobile software is probably unable to use more than a single core, let alone eight, which means that singe-core performance is of the utmost importance to a top-shelf user experience. Many of these "octa-core" solutions sacrifice single-core performance for the insane core-count marketing point.

That being said, it's likely that the high-end custom solutions from Intel, Qualcomm, and even Apple long-term, will top out at four "usable" cores. Apple, in fact, appears to be pushing the envelope on single-core performance and opting for dual-core solutions in its flagship iPhone and iPad products. It's likely that this will continue for at least another generation. But, of course, Apple doesn't need to compete on core count -- the iPhone brand and iOS are what sells these phones.

Foolish bottom line
Will Intel's Merrifield platform gain meaningful traction in phones? Probably not. With a quad-core variant coming shortly after Merrifield, and with the company's integrated LTE part known as "SoFIA LTE" featuring quad Silvermont cores, it seems that Merrifield will be short-lived indeed. Does that mean I expect it to not show up in any designs? No, but the real commercial success duo should be SoFIA LTE at the low end and Broxton from the mid-range up in 2015.

The biggest thing to come out of Silicon Valley in years
If you thought the iPod, the iPhone, and the iPad were amazing, just wait until you see this. One hundred of Apple's top engineers are busy building one in a secret lab. And an ABI Research report predicts 485 million of them could be sold over the next decade. But you can invest in it right now, for just a fraction of the price of Apple stock. Click here to get the full story in this eye-opening new report.

The article Here's Why You Should Cut Intel Some Slack originally appeared on Fool.com.

Ashraf Eassa owns shares of Intel. The Motley Fool recommends Apple and Intel and owns shares of Apple, Imagination Technologies, Intel, and Qualcomm. Try any of our Foolish newsletter services free for 30 days. We Fools don't all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors. The Motley Fool has a disclosure policy.

Copyright © 1995 - 2014 The Motley Fool, LLC. All rights reserved. The Motley Fool has a disclosure policy.


Increase your money and finance knowledge from home

Bonds for Beginners

Learn about fixed income investments.

View Course »

Investor’s Toolbox

Improve your investing savvy with the right financial toolset.

View Course »

Add a Comment

*0 / 3000 Character Maximum