Print 40 comment(s) - last by jihadjoe.. on Jul 24 at 3:51 AM

Mobile demand is driving Intel into the low-end chip market, the last place it wants to be

Analysts had hoped [source] that Intel Corp. (INTC), the world's largest maker of chips for traditional personal computers, would post revenue of $12.9B USD and a profit of $0.40 USD/share ($2.05B USD).  Instead the chipmaker slumped on slow PC sales, posting a net income/profit of $0.39 USD/share ($2.00B USD) on revenue of $12.8B USD.

I. Intel CEO Admits to Mistakes

The miss might seem small, but for investors it was very unsettling.

The sentiment is that while Intel may be profitable for now, the company missed the major direction the market is heading towards: mobile devices.  Profit slid 30 percent in a year and analysts are wondering if that trend will continue, leaving Intel in dire shape in just a couple more years.  Driving this slide is the trend towards cheap tablets (hybrid tablets/laptops) and smartphones which are cannibalizing traditional PC sales.

Brian Krzanich, Intel CEO [Image Source: AP]

Needless to say, it was a touch quarter for Intel's new chief executive, Brian M. Krzanich.  In an earnings call he offered up a mea culpa of sorts, stating:

Intel was slow to respond to the ultramobile PC trends.  The traditional PC market segment is down from our expectations at the beginning of the year, while ultramobile devices like tablets are up...  At the end of the day, the market will go where the market goes.  We’ve not always lived up to the standards we've set for ourselves.

In the call, he promised lower price points for chips and cheaper devices -- particularly on the low end.

II. $150 USD Tablets With Intel Processors? Intel CEO Promises Big

Key to that effort will be the upcoming sixth major release to the Atom platform.  Manufactured on the latest 22 nm node the quad-core tablet-geared Atoms (core: Silvermont; SoC: ValleyView; chipset: Bay Trail) are expected to be shipping in tablets and hybrid notebook/tablets this holiday season.

Mr. Krzanich boldly predicted the following price points:
  • Convertibles (hybrids) : $400 USD
  • Laptops (w/ touch)      : $300 USD
  • Tablets                           : $150-200 USD
The big question is whether these devices will be appealing customers.  Some $200 USD and lower tablets -- namely, Google, Inc.'s Nexus 7 -- have received good reviews.  But current Intel tablets have been priced much higher (for example the Iconia W3 Acer Inc. (TPE:2357) at $379 USD) yet earned worse reviews.

A big question is how low Intel is willing to go on chip pricing.  Intel typically has charged much higher rates per chip than ARM chipmakers, citing the processing power of its chips

Haswell ultrabooks and tablets
Intel would rather customers buy pricier Haswell based tablets/convertibles, but may be forced to turn to budget models. [Image Source: Intel]
But ultimately the CPU is proving less critical to the tablet experience than other components like the touchscreen, battery, and operating system.  A tablet with a less powerful CPU, but a better screen, battery, and OS is often perceived as superior to a tablet with a faster, more expensive CPU but otherwise weaker specs.

This has come to a shock to Intel, which has long put its faith in the philosophy that the fastest chip built on the best process sells.

Intel and its partners will have to make big changes to have any hope of hitting the price points Mr. Krzanich was advertising in the earnings call.  Take Acer, for example.  It would have to cut $172 USD off the current price of the Iconia W3, which has a 1.8GHz dual-core Intel Atom processor.  That's a whopping 47 percent price cut, and one that will surely demand much lower unit prices from Intel.

III. Chromebooks Surge as Intel Struggles to Adjust

If Intel can't hit those prices, it will likely lose more market share.  The NPD Group reports that Google shockingly scored a quarter of the sub-$300 USD laptop market, capturing about 4 to 5 percent of the market, according to Gartner, Inc. (IT).  

Intel does power one of those Chromebooks -- the Acer C710 ($220 USD), which packs a Celeron 847 (1.1 GHz dual-core 32 nm Sandy Bridge).  But the other Chromebook -- Samsung's $250 USD 11.6-inch model -- packs Samsung's in-house 1.7 GHz dual-core Exynos 5250 processor, which uses an instruction set from ARM Holdings Plc (LON:ARM).  Samsung's model was the #1 top notebook seller on, Inc.'s (AMZN) computers and accessories section and had an average of 4.1/5 stars; the Intel-equipped Acer received 3.3/5 stars and was the #8 top notebook on Amazon.

Acer C7 Chromebook
An Intel-powered Samsung Chromebook at $250 USD might sell at least as well as the current ARM model and is possible, given Intel's efforts to woo Samsung.  Intel recently scored a contract to power a variant of the Galaxy Tab 3 10.1, which runs Google's Android OS.

But even if Intel can win the low end tablet/notebook/hybrid market with a mix of Atom-based Android and Windows devices, it faces a catch-22; those sales may cannibalize unit sales of higher end, higher margin chips like Haswell.

Intel has never had a penchant for the low margin, low-end processor business, which it left largely to ARM's chipmaking coalition.  The fact that it's committing so fully to it now, and even contemplating acting as a third-party contract fabricator for ARM chips shows how fearful Intel's leadership is of low-price mobile devices cannibalizing sales of higher-priced traditional PCs.

IV. Process Lead May Be Lost by 2015

The situation could be growing more dire soon as Intel.  Samsung has reportedly committed to delivering a large volume of 14 nm processors to Apple, Inc. (AAPL) by 2015.  Meanwhile, reports indicate that Intel will shelve its own 14 nm (3D) FinFET rollout till 2015.  Unless Intel surprises or Samsung flinches, this would indicate that for the first time in over a decade Intel might no long have a process lead.  That's insult to the injury that the mobile device pickup has dealt Intel.

14 nm product from Intel is reportedly shelved till 2015. [Image Source: Intel]
In the earnings call Mr. Krzanich says that 14 nm will enter mass-production by the end of 2013.  This indicates a late H1 or early H2 launch of the 14 nm Broadwell may be possible.  But even if Intel delivers, Samsung may not be far behind.

Mr. Krzanich insisted in the conference call that the low-priced Atom push wouldn’t cannibalize Haswell device sales, instead creating new business.  But the reality is that whether it does or not he has no choice -- as he says, "the market will go where the market goes."  

A year ago Intel made $2.8B USD in profit.  This year it made nearly 30 percent less -- $2.0B USD.  With PC sales at an all time low, Intel is diving into the shallow waters of low-budget bargain chips -- the last place it wanted to be.

Sources: Intel, NYTimes [earnings estimates]

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

SSDs is where Intel should be focusing on
By someguy743 on 7/18/2013 2:56:57 PM , Rating: 2
I don't think higher and higher performing CPUs are really needed as much as they used to be. The bottleneck was definitely my HARD DRIVE on my Ivy Bridge i7 computer. That is, until I bought myself a new Samsung 840 Pro SSD. It improved the performance of my PC dramatically. Night and day difference.

I would like to see Intel start cranking out 14nm (or smaller) SSDs that are ultra fast and store 1 terabyte or more of data ... all for a reasonable price of maybe $200 or less. That would be awesome. Intel would made piles and piles of money.

The latest and greatest SSDs need to be super durable and last at least 5 years too. Maybe Intel could have durable enterprise grade SLC NAND SSDs for consumers in a few years that are 1 TB and only cost $200 or less. That's my dream anyway.

I can't wait until I never have to use another traditional hard drive in my life at work or anywhere else. That's how much I love my new Samsung SSD. Out with HDDs and in with the new SSDs !!

By someguy743 on 7/18/2013 3:55:21 PM , Rating: 2
This is the kind of product that I would like to see Intel use their deep knowledge and experience on. It just came out in the last few days.

Samsung XS1715 SSD ... that uses high-speed NVM Express (NVMe).,2817,2421965,

They have capacities as large as 1.6 TB and they are SIX TIMES as fast as today's fastest SSDs! 3,000 MB/second. My Samsung 840 Pro does about 500 MB/s now. I'm sure it is insanely expensive for almost all of us consumers though.

I think Intel needs to bring some serious competition to these SSD markets. Right now it's looking like Samsung is dominating.

If Intel can dramatically bring down the price of products like this new Samsung XS1715 SSD they will make major buckeroos for years to come and get lots of fanboy love from IT guys and everyone else.

By Ahnilated on 7/19/2013 12:03:30 AM , Rating: 2
Have you read anything about the smaller processes making the writes to an SSD burn them out even faster? This could be a massive problem if your data gets destroyed in 3 yrs because of normal use. Granted, everyone will say, well you should be backing up your software. That doesn't mean every 3 years I want to have to buy new SSD's just to have a decent speed system. It seems like we are allowing companies to sell us stuff that they know will fail in 3 yrs so they can sell us more stuff, IE planned obsolescence.

By Azure Sky on 7/22/2013 11:24:45 PM , Rating: 2
I think your best bet would be hybrid style drives, as somebody else points out, smaller process nodes are currently leading to quicker burnout of flash memory, they havent fond a fix yet.

a proper hybrid drive with the right flash/cache/hdd ratios could really kick some arse.

large capacity, fast speeds, lower fail rates then pure flash(flash even ent class can be burned out pretty quickly by high write cycles, seen a few intel/ocz/samsung/adata ent class drives get smoked in under 6 months...the fix is to use a software like fancycache to delay writes and avoid writes that dont need made, this could be done ON DRIVE using large on drive ram caches, using a capacitor setup that gives the drive time to flush data in ram to flash could fix the potential of data loss.

IMHO, a drive designed with say 4gb ddr3, 16-64gb ssd(blazing fast high durability flash) then a 2-4gb hdd for storing less used data, setup so that the ssd caches the most used clusters of the hdd and the ram caches writes and the most accessed data from flash...could make for insane drive speeds without the need for high density flash memory that frys itself quickly.

"I mean, if you wanna break down someone's door, why don't you start with AT&T, for God sakes? They make your amazing phone unusable as a phone!" -- Jon Stewart on Apple and the iPhone

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki