Celeron Inside means Confusion Outside for the Ultrabook

Posted on 21 May 2012 by

celeron

There was a brief report a few days ago that hinted at new Celeron CPUs being offered for Ultrabook products. I don’t believe it. Not only does this source appear on my personal ‘blacklist’ but there are good reasons why this report is wrong.

Update: The news item was updated with a response from Intel. We were right, Celeron CPUs are not going into Ultrabooks.

Most people following, reporting and selling Ultrabooks getting to the point where they are getting a feel for how an Ultrabook performs and what features it will offer. Identity Protection Technology, Intel Insider, Quick Sync Video, Clear Video and a level of processing power that brings swift productivity. Intel hasn’t laid any of this out in stone (and already have the Core i3 processors that strip out the useful Turbo features) so there’s scope for changes but there’s something else that doesn’t make sense here. Ultra low voltage processors are among the highest quality products to come from the production lines and that’s the reason they can be set-up to be so efficient, with such low TDPs. Celeron processors, on the other hand, might be cut from the same wafer using the same process but they are far from the highest quality parts. Lower TDP versions of these processor are made by cutting out features, lowering clocks, and disabling cores. That’s not what the Ultrabook is about. Do that and you’re in cheap laptop territory and confusion for customers. Total confusion for anyone reporting on the Ultrabook and bringing news and reviews to customers.

Regardless of any technical or idealistic wishes, it’s something that Intel could do in order to cut $50-100 from the price of the CPU. The products could sit in the same chassis as the Ultrabook-focused products and be called ultrathins. In a way, it makes some sense to offer manufacturers an Intel solution for the products that AMD Trinity is aiming at but there’s such a huge risk of diluting what the Ultrabook is that I can’t see Intel doing it. I truly hope they don’t. Or does a $599 Celeron-based Toshiba Z830 appeal to you?

Categorized | Opinion

Tags : , ,

  • Tsuki

    AMD has a stronghold in the space between Atom and ULV Corei in the form of Zacate, Trinity will only push further into Intel territory. Intel doesn’t like this. Atom isn’t able to compete with E series APU’s much less 17w Trinity, thus Intel’s doesn’t really have a choice but to pump out cheaper ULV parts.

    Intel has shown that they are willing to dilute ultrabook’s premium-ness with budget materials, which imho is a much worse offense than low end CPU’s.

  • Adam

    Tsuki, I think you’ve nailed it. Intel realizes that they need to compete on price or they’re in a world of hurt but investors are just waiting to see if Intel attempt to compete with ARM by dropping profit margins which would spell stock price disaster for Intel. This is a smart move by Intel but I’m not sure its going to be enough. Intel actually has to compete on perf+price and in an app and mobile-centric world perf matters less and less and efficiency, mobility, and price more and more.

    This is a logical step for them but I don’t think it’s enough. They’ll need more lower profit mobility parts going forward.

    A lower price point is what people have been demanding in Ultrabooks and rightfully so when comparing to the Core2-based ULV CPUs efficiency and prices…

    Ultrabook officially means both everything and nothing now. Except “thin, light, some sort of SSDish drive somewhere, and Intel”.

    Adam

    • James

      Let’s not confuse the ATOM with Intel’s Core i-Series. The ATOM has always been a low margin product for Intel and there’s little question they’ll be pushing it to be cost competitive with ARM.

      Cedar Trail already nearly halved the costs of the ATOM for Intel already. The Pine Trail N550/N570 for example was rated for about $86 and the Cedar Trail N2600 and N2800 were rated at $42 to $47 respectively when they first came out.

      While top of the line ARM chips are tipping over the $20 mark now. So the divide isn’t insurmountable for Intel given that they’re going towards cost saving SoC designs and have another die shrink due next year to 22nm to further help reduce costs for them before they even have to think about touching their margins.

      They probably have more to worry about the added cost of Windows 8 for Intel based devices than their hardware costs.

      Also don’t forget that the Medfield shows Intel can get power consumption down to ARM range already. Something AMD may not even attempt outside possible tablet targeted offerings.

      I agree though that AMD stands to leverage itself well in the space between what Intel ATOM’s offer and Intel’s higher end Core i-Series systems, and Intel could be in trouble unless they can get their Core i-Series costs down for Ultrabooks.

      But… Mind Haswell is coming next year and it’ll also bring in a more SoC/MCM design to help reduce costs and will offer a better range of performance and power efficiency than Ivy Bridge.

      Though AMD will also be coming out with their next gen offerings next year too. So we’ll see how they’ll compare but unless one of them stumbles it looks to be holding status quo, but at least we’ll have more choices made available by then.

  • Adam

    As the ARM CEO has recently said “We’ll be bigger players in PCs than Intel will be in smart phones.” -Agreed completely based upon current trajectory. If Intel can’t compete in the high efficiency, low profit space, they’re in a world of hurt. They HAVE the backwards compatibility story but it doesn’t matter much in the app-centric world if they can’t or are unwilling to provide high efficiency, low profit devices that continue to perform slightly better than the other guys.

    Adam

    • James

      CEO’s often try to spin things in their company’s favor. Doesn’t mean it’s even remotely accurate.

      Take the so called current trajectory… do you for a second factor that ARM was never intentionally designed for the PC market?

      Often niche centric products, rapid end of life cycles, limited to no legacy support where every device and app is almost treated as disposable, rampant hardware fragmentation, closed driver support, limited to no hardware flexibility/upgrade-ability, and limited performance scaling that’s is often optimized for lower performance to get better power efficiency and lower costs.

      Many of those are typical ARM strengths that allow them to be quickly optimized for a given usage need and minimize costs but they go against what has typically been strengths in the PC market.

      Legacy support, hardware standardization, ease of configuration flexibility/upgrade-ability, long periods of support, more open drivers, compatibility with a wider range of software, higher performance scaling and more flexible usages.

      Mind it’s generally easier to lower performance and increase power efficiency than it is to raise performance and not sacrifice something in the process.

      ARM is still a 32bit processor in fact and it’ll take them a few years at minimum to move past their present next gen offerings and start developing a range of full 64bit offerings.

      Really, ARM needs to make a lot of fundamental changes to really compete in the traditional PC market.

      Budget wise ARM FABs have about as much revenue stream as Intel’s, per manufacturer, but they don’t exactly work together as much as compete and doesn’t mean they’ll have it as easy as Intel.

      TSMC for example has been in the news a lot lately with issues ranging from having to invest more money to fix their 28nm process, reducing the range of their 20nm process to only support one flavor (no more low to high end offerings), and stating doubts about whether they can even go 14nm and may opt for 18/16nm half nodes instead.

      While Intel is not only on course for 22nm, but 14nm, and more recently reported down to even smaller single digit nm and all on schedule so far.

      Among other examples like the 32nm FAB update Apple recently used to update their A5 employs HK+MG that Intel started using with their 45nm FAB but ARM didn’t start using until they went 32nm.

      So it’s not like Intel won’t have any leverage to work in its favor.

      Sure, ARM will get into a lot of devices and probably far more than any x86, but more often than not they’ll more likely be used for specialized purposes than all purpose PC’s.

      Especially if both Intel and AMD manages to strengthen their positions with similarly affordable hardware but retain the scalability and other strengths that have traditionally helped them succeed in the PC market.

      Though, given that’s it’s mainly just AMD and Intel in the PC market, it’s sort of true that ARM will do better considering that Intel will have to compete with literally thousands in the ARM market and will have a much harder time getting a significant portion of that pie but that’s a different matter from whether they will do well or not.

  • DavidC1

    Well…

    The same report you are criticizing says this:

    Intel has responded saying that “the report suggesting Celeron powered ultrabooks would come to market is false. In order to deliver ultrabook devices that wake quickly, and offer a minimum of five hours of battery life, Intel requires all ultrabooks be powered by an Intel Core processor.”

    So they can just market these as typical thin and light laptops.

    • Chippy

      Amazing that Digitimes updated that story.
      I’m glad we’ve got that out of the way.

Trending UMPCs

Acer Aspire E11 ES1
11.6" Intel Celeron N2840
Lenovo ThinkPad P40
14.0" Intel Core i7 5500U
Asus E202
11.6" Intel Pentium N3700
Lenovo Ideapad Flex 10
10.1" Intel Celeron N2806
HP Spectre x2
12.0" Intel Core m3 6Y30
Lenovo Thinkpad Yoga 260
12.5" Intel Core i3 6100U
HP Pro Slate 8
7.9" Qualcomm Snapdragon 800
Acer Aspire Switch 10
10.1" Intel Atom Z3745
HP Pavilion X2 10
10.1" Intel Atom Z3745
Teclast x2 pro
11.6" Intel Core M 5Y10c

Recommended Reading