Intel's Haswell launch may take backseat to Atom vs. ARM

Intel's Haswell launch may take backseat to Atom vs. ARM

Summary: There's an increasingly vocal chorus of analysts who believe that Intel can solve for ARM and ultimately pass that architecture in mobile computing performance.


Intel is widely expected to launch its latest PC processors at Computex, which will outline an array of new Windows 8 devices, but the real game may revolve around Atom's future roadmap and mobility.

Naysayers will note that the chip giant's prospects against the ARM architecture aren't good. These naysayers are also likely to add that Intel's Haswell-based chips are designed for a PC industry that is getting thumped by tablets.

The argument against Intel goes like this:

  1. ARM, a low power processor architecture, will dominate as mobile computing (smartphones and tablets) routs PC sales in the years ahead.
  2. Intel's manufacturing prowess and focus on performance is a waste in an era of good enough computing.
  3. Intel doesn't seem to have a mobile answer that's been given a thumbs up by the market.
  4. New Intel CEO Brian Krzanich is likely to stay the course so the chip giant may never solve for ARM.

Folks following the semiconductor industry can recite those arguments in their sleep. To many folks, Qualcomm is the new Intel. Of course, Intel begs to differ with that opinion.

Now there's an increasingly vocal chorus of analysts who believe that Intel can solve for ARM and ultimately pass that architecture in mobile computing. In other words, Intel's recent design wins with Samsung and Asus at ARM's expense are just the beginning.

These analysts are largely backing Intel based on the company's next-gen Atom roadmap dubbed Silvermont. Intel presented its roadmap earlier this month and plans to improve Atom every year. Silvermont will have a burst frequency to handle various use cases.



In a massive report this week, Wells Fargo analyst David Wong outlined the case for Intel. The bottom line is that Intel should be viewed as a universal processor company, not just one tethered to the PC industry.

Here's a look at Wong's money quotes:

We believe that Intel’s strong technology position could give Intel’s tablet and smartphone products a competitive edge over ARM-based chips in 2014 and 2015, and beyond.


Intel management has "also put in place what we consider to be an appropriate strategy to penetrate the tablet and smartphone space."


The investment community is appropriately enthusiastic about the wave of mobile computing that is sweeping the globe. However, we think that many investors incorrectly conclude that the growth potential associated with this is limited to smartphones and tablets driven by ARM-based processors.

How will Intel back up Wong's quotes? Wong's bet is that Intel can capture more than 10 percent of the smartphone market and 30 percent of the non-Apple tablet market. Wong also noted that Apple could use Intel's Atom chip at some point in the future for the iPad.

According to Wong, Intel's roadmap indicates that Atom will surpass ARM in power and performance in the future. However, Intel's Atom gains won't move the revenue needle too much relative to other businesses.

Morgan Stanley analyst Joseph Moore also noted that Intel's Atom roadmap could ding ARM's momentum. Moore said:

Atom could be a game changer in Windows 8 tablets. In fact, we would ask the question, does Microsoft need to keep investing in Windows on ARM? The Windows on ARM initiative was born of Intel’s failure to produce compelling performance per watt. With significantly improved CPU performance, and consumers’ preferences for legacy PC applications that run best on x86, we think Intel can achieve leading share in Win8 tablets and some footing in Android. We estimate Intel will ship 50M Atom chips for tablets, the majority being Win8, by 2015.



The difference this time for Intel is that its Atom roadmap is designed to optimize power per watt from the ground up.

If all goes well, Intel's Atom platform — and tablets — that will cannibalize the ultrabooks powered by the company's i-series being shown off at Computex. That "if" is a big one, but more analysts are giving Intel the benefit of the doubt and thinking it can turn back ARM's momentum.

Topics: Processors, Hardware, Intel, Mobility, Smartphones, Tablets

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Anyone that thinks..

    Intel can't compete is either a fool or has not done their homework.
    • Maybe they've got an Axe to Grind.

      That, or they're AMD fan-boys.
    • This is not news to anyone following the chip sector.

      Intel kills arm in perf and manufacturing process tech and with haswell and silvermont will also beat arm on power. When airmont comes out next year after the 14nm shrink the gap will widen even more and on top of that Intel can produce these chips at lower cost with their new large wafer high yields. Any smartphone or tablet OEM not already planning on dropping arm and moving to Intel will find themselves with non competitive product that will be WalMart bargain bin fodder and then not move at all. Intels plan is solid and it's already "the course" that Brian Krzanich is staying.
      Johnny Vegas
    • I agree

      Just picked up an HP Envy X2. Dual Core x86, fanless tablet/netbook the comparable to a Mac Air and ipad. 12 hours battery and runs everything x86 and Android I need! Makes my Dual Core Sony Tablet S Android look dated.
    • Those same fools are the ones

      that say MS dead which is a joke. I have Atom in my Windows 8 Samsung Ativ and it's pretty impressive. I found it to be faster then my Surface RT in most cases and I get around 9hrs of battery life.
    • Yes, we went down this road, in a similar fashoin before.

      Ive mentioned in the past, several times that there was a time back when it was Pentium IV vs AMD in the new 64 bit versions and AMD was clobbering Intel and vast numbers of people were predicting Intel was about to crumble. Even when the Conroe chips were released at the time the nay sayers wouldnt give. Many continued to make every excuse under the sun as to why the cheaper and more powerful Conroe chips would not beat AMD.

      But it did.

      Now here we go again. Isnt it just a little early to predict horrible things for Intel based on such claims like we see now.

      I just cant see how anyone could ever count Intel as down and out at all when Intel just keeps hammering at the issue and they continue to improve there position substantivly every time.

      Why cant we just wait a little for some real and serious results for a bit more of the longer term.
  • Wrong focus

    Intel is focusing on the wrong thing here. I have an arm tablet and one based on the newer atom. The atom based tablet already has great battery life. However the arm tablet has much better graphics performance. This is an area that has always plagued the atom processors. Intel needs to improve the graphics processors on their SoC's.
    • And the Intel line in general

      Apple has historically chosen chips that have a graphics edge, even to the detriment of the computing power. Back in the 90's a math co-processor was not on most of the Apple computers, placing them at a distinct disadvantage for business work. Intel on the other hand, had focused on the computing power, integrating the math co-pro as early as possible. It really reflects the differing targets of companies.
    • HP Mini 311

      The graphics performance on my atom powered HP Mini 311 is good, however it is not hobbled with Intel GMA graphics, but has some NVidia Ion action.
    • The Current Atom uses Ancient Microarchitecture

      Clover Trail the current Atom is the original Atom microarchitecture from 5 years ago and 32nm lithography. The Atom going in to production later this year is a completely new microarchitecture which will debut on Intel's latest year old 22nm FinFET transistor. aka Silvermont.

      Intel is well aware of the graphics issue and has significantly increased their graphic R&D budget the past couple of years. Expect to also see significant progress in graphics later this year.
    • The already are but you make a good point.

      In my opinion ARM doesn't need to worry massively. I think it will take a while for Intel to pass ARM on performance per watt or even price. But at the same time I think all this means that ARM won't be taking over in the server arena any time soon.

      Intel has been making really good progress in the GPU area. They need to do more or they will continue to get killed in the tablet area.
  • Banning non-touch based PCs

    I think MS should consider restricting all new Windows 8 PC systems sold with displays, be touch based. MS could impose a deadline like the middle of 2014 for the restriction to kick in. Selling Windows 8 systems without touch, is kind of like selling Windows 95 systems without mice - where users are given the ability move the cursor on the screen with the arrow keys. OEMs need to move away from non-touch systems, as this is hampering sales; but OEMs are not going to do this on their own. For Intel, MS, and OEMs to do well, I believe imposing this restriction is very important.
    P. Douglas
    • Off topic, but I agree.

      W8 is impossible to use without getting uber-frustrated if you don't have a touch screen. When I check out new laptops and I see W8 running, the first thing I do is touch the screen. If it doesn't respond, I move on.
      Jacob VanWagoner
      • Touch screen price and availability be damned?

        Especially the latter?

        And what about the consumers? Do they have a say?

        That will kill the PC market for sure.
        • @Jacob VanWagoner a Consumer With a Touch Screen...

          echos the voice of most all Win8 users with a touch screen.

          You can resist all you want, some day you will eat your words when you eventually (inevitably) buy a touch screen.
          • Bull

            Care to provide citations? In fact, several surveys have shown that most people with touch screen devices do not use the touch screen much.
            As to the argument that non-touch PCs are killing Windows 8 sales, first, Win8 sales are just fine, at least in to the channel. More importantly, the sales data of devices WITH touch, or rather the lack of sales, indicates otherwise.
          • I have a touch screen

            windows 8 laptop and I rarely use touch. In fact I bought a mouse along with the laptop.
          • Really?

            I'd love to watch you work on a CAD drawing with a touch screen, I could use the laughs.
      • Off topic: yes and no

        I admit my comment was off topic in the sense that the article was about Intel chips, and my suggestion was about Windows 8. However, in the larger sense, it wasn't. The author was alluding to Intel responding to the ARM threat from mobile computing. I believe for the Windows 8 ecosystem (which includes Intel) to counter its ARM based competition, it is really important for the former to fully embrace touch, instead of producing half-baked solutions the market really doesn't want. Trying to counter ARM based touch experiences with a mix of non-touch and touch experiences around Windows 8 is not working. People overwhelmingly like Windows 8 touch experiences, and either outright hate or sort of like Windows 8 non-touch experiences. Just get rid of Windows 8 non-touch experiences in the form of new systems, and there is a good chance that Windows 8 systems will pick up in growth fairly quickly, and will be able to counter its ARM based competition.
        P. Douglas
        • And yet, sales of touch-based Win8 systems have been disappointing.