Intel recently announced plans to layoff 11 percent of its workforce as the company shifts its focus from PCs to the cloud… now we’re starting to get an idea of what that will mean for the company’s line of chips for personal computers: the low-power, entry-level Atom chip family is being phased out.

Those are the chips commonly used in cheap, low-power tablets, notebooks, and 2-in-1 devices.

Cherry Trail chips have a bit of life left in them, but the planned Broxton chips that were set to replace them have been cancelled. And Intel does still plan to launch new low-power Celeron and Pentium chips based on the Apollo Lake platform soon.

Update: It turns out the Atom brand is alive and well… but it’s not being used for smartphone, tablet, or notebook chips anymore. Instead, Atom chips are powering new embedded and IoT-oriented devices like Intel’s new Joule system-on-a-module development platform.

intel

As Forbes reports, Intel is also cancelling its SoFIA chips designed for smartphones and other  lower-power devices.

PC World, meanwhile, reports that Intel is also going to end production of the existing Intel Atom Cherry trail chips.

Update: Intel tells me the company “will continue to ship Cherry Trail Atom x5 and x7” chips.

broxton

Honestly, neither of these things should come as a surprise if you’ve been following the PC space in recent years. Device makers are practically giving away tablets and low-cost laptops with Intel Atom processors.

There’s not much profit to be made in the space. Microsoft even gives away its Windows software to makers of small tablets in an effort to just get more users in the space… and to make its operating system somewhat competitive with Android, which is also free for device makers to license.

So while it’s been a rather exciting time for customers looking for a decent laptop for under $200, I’m not surprised to see Intel decide to focus on more profitable areas.

atom

Does this mean that low-power laptops and tablets are dead?

Nope. An Intel spokesperson tells Liliputing that the company will “continue to work with OEMs to develop new 2-in-1s based on Apollo Lake” as well as Intel Core M chips. So while Atom chips will eventually be phased out, we’ll continue to see low-power tablets that offer somewhat better performance… most likely at a higher cost.

And the company will continue to push its Core M chips for higher-priced, higher-performance devices. These chips don’t use much more electricity than an Atom processor, but offer significantly better CPU and graphics performance. Computers with Core M processors often sell for twice the price (or more) of a similar model with an Atom chip, though.

For instance, the 2016 Intel Compute Stick mini PC with an Intel Atom Cherry Trail processor is priced at about $140, while a model with a Core M processor sells for closer to $400.

core m

It’ll be interesting to see if Intel’s move away from the chips used in sub-$200 devices leaves an opening for rivals like AMD or some of the many chip makers using ARM designs.

While you can’t run the full Windows 10 operating system on an ARM chip, Chrome OS and many Linux-based operating systems work just fine. And Intel’s decision to get out of the Atom space will have little to no impact on some of the most popular tablets including Apple’s iPad lineup, Amazon’s Fire tablets, and Android tablets from Samsung, Asus, and others.

I’ve seen some headlines suggesting that Intel is giving up on tablets altogether, but ending the Atom line of chips really just means the company is giving up on cheap tablets. We’ll continue to see high-end tablets and 2-in-1 devices like the Microsoft Surface Pro.

surface pro 4

What this means for entry-level Surface tablets remains to be seen (the $499 Microsoft Surface 3 has an Intel Atom x7 Cherry Trail processor).

Something else to keep in mind: Intel’s Atom chips were kind of synonymous with netbooks for a few years before they started showing up in mini-desktops (remember nettops?), tablets, and larger notebooks. But the laptop that was arguably the first consumer netbook didn’t have an Intel chip. The Asus Eee PC 701 launched in 2007, it had an Intel Celeron processor because Intel hadn’t released an Atom chip yet.

But the Eee PC was a tiny laptop priced at under $400 at a time when that was unheard of. By using available low-cost components, Asus managed to produce an affordable portable. I suspect if device makers still see demand in this space, they’ll find a way to continue doing that, whether Intel offers chips specifically designed for the purpose or not.

Support Liliputing

Liliputing's primary sources of revenue are advertising and affiliate links (if you click the "Shop" button at the top of the page and buy something on Amazon, for example, we'll get a small commission).

But there are several ways you can support the site directly even if you're using an ad blocker* and hate online shopping.

Contribute to our Patreon campaign

or...

Contribute via PayPal

* If you are using an ad blocker like uBlock Origin and seeing a pop-up message at the bottom of the screen, we have a guide that may help you disable it.

Subscribe to Liliputing via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 9,547 other subscribers

111 replies on “Intel is killing off low-power Atom chips (for smartphones and PCs)”

  1. Bad decision by Intel for sure. This could have been a goldmine without any real competitor for the long foreseeable future.

  2. Hmmm well I guess that’s the end of the Microsoft Surface then. Would have been nice to see what updates they would of made to it this year. If the alternatives cost more I cant see Microsoft continuing with the model, they’ll just stick to the Surface Pro instead.

  3. Rip the entire chinese tablet business?

    I liked my 80 dollar windows tablets 🙁

  4. Its a pity, but not suprising all of atoms markets are struggling

    50 quid tablets with core duo level processors were clearly a big loss leader by Microsoft and Intel and attempt to squeeze out android and ipads, i think they have failed because of the lack of touch friendly software. Its a pity because any device I did use was fantastic for browsing the web and retro gaming. Android devices that used the atom tended to have a few issues here and there which led me to believe the mainstream arm chips were a better choice.

    Cellphones likewise is a tough market, I think atoms android performance wasn’t enough for flagship phones and again I think the hassle of running android on x86 caused headaches in the background.

    The whole Atom thing was time limited from the start, they went from lacklusture single core chips that were never viable for anything more than a netbook to being cheap relatively powerful chips that could eat into celerons market share. If you are getting ahead of the competition theres no point in churning out these chips for half nothing and canbalizing your own market share. Its a pity I’ll miss the cheap tablets and stick PC’s.

  5. Intel is on their way down just like Apple! …….and SAMSUNG is KILLING BOTH OF THEM… in Long Term Growth. They are gaining market share Apple is now losing and just Samsung Semiconductor Business is in the Process of SMACKING DOWN INTEL BY 2017… as the #1 Chip Maker in Profits and Revenue…. and smoking them on innovation and Fabrication Process Leadership!

  6. This is a great shame for me from the perspective of the TV box market.

    Until Atom came along we had the usual stream of new ARM based hardware backed up by continuous crappy SDK’s and ROMs from the makers that meant that you could never get out of the devices what the advertising promised.

    I’ve lost count of how many sticks and boxes that I have bought and the hundreds of custom ROM’s to hope to get even close to the stated expectation.

    Then along came Atom and in particular Cherry Trail and in one go I was able to do all the things that I needed to without so much as a custom anything and far less support required from the vendors.

    Back to crappy ARM in the future by the looks of things.

    Unless AMD can spot a hole in the market.

    1. Unfortunately, AMD pretty much gave up on x86 for the embedded market, instead they’ve adopted a ARM version of their APU and seem focused on the server market. So only look for something in the low cost laptop on up range from them with x86…

      Though, some OEMs may opt for Apollo Lake to still make some devices in this category… It’s still possible to go fan-less with a large heat sink and/or Intel may offer a lower powered version of it… Mind, it’s the same Goldmont architecture as Broxton would have used, just optimized for the Celeron/Pentium range of devices.

      But otherwise, yes… ARM will be by far the dominantly offered option in this category range going forward…

  7. What does this mean about the future of Hololens since it’s based on the atomx5?

    1. Not much right now, they’re not discontinuing the x5… They’re just not going to offer a update to it… So it will only be a issue whenever they plan to upgrade the Hololens…

      While we’ll see if Apollo Lake will be offered in a wider range of devices to compensate or not… It’s still the same Goldmont architecture as Broxton would have used, so there’s some flexibility there…

  8. Dang. There goes my hopes for someday buying an x86 smartphone with Atom and a full Windows version with desktop. Also really wanted an Atom-powered NAS and one of the mini gaming handhelds that should come out this year. Already have a few older Atom netbooks and this year’s Atom Compute Stick, and they have always been what I’ve expected Atom to be.. low-cost, low-power and excellent value for sub-10″ devices.
    Hopefully, Core M will eventually replace this segment, rather than Celeron and Pentium, which I’ve always seen as poor value and much less energy-efficient for these smaller devices. The Pentium brand was pretty cool 20 years ago.. not so much today, and I’ve never thought highly of the Celerons (even less so than the Atom branding, for whatever reasons).

    1. I feel you! But truth be told there are no Pentiums and Celerons since the Core era, only rebranded low cost versions of the others. A higher specced Atom would get a Celeron or Pentium branding and so does a low-end Core i-something product. Personally what bugged me for ages that you could get a Celeron that would be slower than your average Atom, but there were Celerons that run circles around them being almost (but not quite) an i3, and there were no way to tell from the modell number what you got under that Celeron/Pentium name until you checked CPU-boss or something.

    2. Right, exactly. Atom was Intel’s and Microsoft’s chance to change the cell phone and low powered device space. What Intel did was squeeze AMD out of this space by lowering the cost of the chips when the competition would have created a price equilibrium that wasn’t going to be artificial.

      Now what are we left with… ARM… and ARM wipes out 20 years of application development in the PC space. Yes yes… charge ahead into the future… I get that. But you’ve got businesses everywhere with old server 2003 compatible drivers for their hardware, and ARM will not interface with that stuff without forcing manufacturers to support their EOL products.

      Compatibility is why we would like the x86 smartphone at a reasonable price.

    3. Zenfone 2 has 4gb ram and Intel x86 atom. Primarily runs android but you can run windows 8.

      Alternatively teclast 98 4g dual boots android and windows 10

    4. None? Very simple.. to have one unified device to carry around. A smartphone that I can pull out of my pocket at work and plug into a monitor or dock w/KVM (or use easily enough via a pull-out stylus) and have my full desktop configured as I need and ready to go. All the same usage scenarios that I use my compute stick, netbook, etc for.. but with the ability to boot into Android or Window Mobile to use as a regular smartphone (ability to accept calls inside full Windows would be another option). I use a few applications that will never be able to accomplish the same tasks with a crappy scaled-down phone app version.

      It’s the exact same idea as the upcoming Microsoft Continuum.. but instead of offering a “PC-like experience”, it would offer a real PC experience. Same thing as the 7″ Windows tablets that are now available, but with a phone radio added in and a little smaller. It could even have a keyboard like the tiny Fujitsu UH900, and several of the other pocketable PCs of long ago (though a slide-out keyboard like the old Nokia N900 would be nice). Dunno.. I just see it as a further evolution of the phone (the same way that DAP, camera and nav functionality were eventually added, rather than lugging around multiple dedicated devices).

      Why would I want such a crazy thing? Perhaps it’s because many of us here like gadgets and small computers. Liliputing still means “very small computing”, right?

      1. Seems like a decent enough idea, but really isn’t a solution at all for my particular needs/wants. But, if this device combo works for you, then I’m glad that such a thing might eventually exist, and I won’t argue up and down about how shit I personally think it is. If that’s what you want, then great.

        I don’t want a laptop shell, I don’t want paired-down Android mobile apps.. just something that can function daily as a smartphone, and then boot into a Windows desktop OS to run full-featured x86 applications when needed (via any connected TV/monitor, or using the built-in screen in a pinch).

        Would be awesome if Microsoft made Windows 10 and Windows Mobile as such a hybrid OS for x86 smartphones, but that may never happen. Full Windows desktops on sub-7″ screens existed back in the UMPC days and people used them productively (never wanted to spend the $2k+ for them back then, but such a thing could be done for a fraction of the price today).

        The whole idea might be terrible for most and will likely never happen at this point in time anyway.. it’s just something I’ve always wanted since the early PocketPC era. I’m not here to sell you on the idea for whatever sliced bread crowdfunding campaign of the week comes next, so I honestly don’t care at all what you think about it, sorry.

        1. So, you picked a random comment from an article several months old just to be an ill-tempered dick and argue for no reason whatsoever about what I *want*, without even attempting to comprehend anything I’ve ever said? Get bent with whatever crusade you’re on, as I really don’t care what you think about anything.

  9. Can anyone clarify how this new relates to Apollo Lake? Is Apollo Lake cancelled as well? I thought that was targeted towards some entry level PCs and laptops. Brad wrote about this very recently, so I’m curious the status.

    1. The news is that Intel is phasing out of the Smartphone SoC market in favor of their more traditional markets… Since Apollo Lake serves the more traditional markets it will still be coming to market but the Mobile centric Broxton and updates for SoFIA will not…

      There is no news on whether this will be a permanent change but it seems for the foreseeable future that Intel won’t be using the ATOM specific branding but that doesn’t mean ATOMs that serve other parts of the market won’t continue… Though, it’s likely they will go through a major redesign and future branding may be completely different going forward as the primary focus may no longer be towards the mobile market and/or they may merge the architecture with the Core series for a more scalable solution…

      Though, it should be noted that Apollo Lake is suppose to be a cheaper solution than Braswell was and the lower BOM, a claimed up to 7% lower, is suppose to take up less total space, and better power efficiency should allow it to cover at least some of the market range that Cherry Trail now covers as well… at least the upper range Cherry Trail…

      1. Thank you sir! To me you are one of the most knowledgeable people on this subject and I appreciate your insight. Thanks!

  10. Windows 10 works fine on ARM, this is why Microsoft is telling everyone to develop Windows Universal Apps, those work perfectly on ARM, on Qualcomm-powered Windows Continuum and more.

    1. And finally we will have a common OS platform to run benchmarks on both ARM and x86/X64 hardware. :3

  11. shit so low cost windows tablet will soon disappear? I have the Chuwi vi8 and I find it pretty good for its use.

    1. They can still make devices below $200, say around $169… We just won’t see anything lower than that… So no more $99 PC tablets…

      Apollo Lake is designed to provide OEMs lower BOM costs, takes up less space than Braswell did, offer better power efficiency, etc.

      So at the very least, they’ll still be pushing Cloudbooks and we are likely to still see pretty cheap China tablets…

  12. The real cause for Intel’s problems lies with programmers who have failed to use up more processing cycles on the PC in any way that’s meaningful to a broader audience. (this could be because everybody is just wasting time on garbage like Fbook, Twatter and Blabbergram instead of any meaningful work)

    Even with a near 5 year old i7 based computer, there is still too little reason to upgrade to Skylake, even though that’s a really good chip with many improvements in the chipset.

    Even for gamers this is so. Nvidia is more likely to drive Intel sales than Intel is.

    When Pascal GPUs become available, I’ll upgrade to Skylake mainly as an extra but non-essential upgrade to Skylake and the benefits in the new chipset’s memory, interface and bus technology improvements.

    Sure, reducing the electric bill is nice too, but unless its a laptop, its not going to turn heads.

  13. Good riddance to Atom… it was always too gimped to do anything with it at a reasonable clip.

    Saving the $50 or $100 on street prices (not unrealistic MSRPs) for an Atom device has really not been worth it for anyone but poor children in the 3rd world. These can be served by better performing, yet cheaper ARM chips.

    There won’t be a single thing that I’ll miss about the neutered Atom line.

  14. They aren’t gonna stop making smartphone chips the problem is there’s no money to be made anymore in LTE smartphones so they’re not going to develop chips for that but theyll back with a vengeance with 5G phones

  15. At first i was a bit angry about this, but to be honest, i will probably never buy another Atom powered product anyways. Perhaps a Compute Stick or something at best.

    If the Celeron and Pentium lines take over that segment, so much the better.

    If AMD steps into that segment with the Zen architecture, I will be thrilled

    1. Agreed. I too have shied away from Atom products since I got burned with the N450 back in the netbook days. Overhyped and underwhelming. Perhaps more folk than I remember that, and avoid the product name.

      1. Well Atom products have come a long way since the days of the N450. In those days, the Atom was hardly capable of running Windows XP.

        Nowadays, even the lowest end Atom with 2gb RAM can run Windows 10 and light software reasonably well.

  16. Poor Microsoft just can’t catch a break. Even while their current attempt at cell phones is dying on the vine their work with Win 10 is beginning to bear fruit. More and more major apps are announcing Universal Windows apps. That means in another year or two Microsoft could have again come at cell phones using those Universal apps and had its best shot yet at the necessary mobile land grab it needs for the future.
    Now this pulls the rug out from under them I think. That or they have a lot more work to do getting Windows to actually run as well on ARM as it does on Intel.

    1. First, this news has nothing to do with MS… Who already put a divide between mobile and PC platforms when they launched Windows 10 and is why they’re pushing their Windows 10 Mobile and not just the Windows 10 desktop OS…

      Universal Apps are also hardware agnostic, it’s one of the reasons why they’re called Universal!

      Aside from bridged apps there shouldn’t be a problem for MS as they’re already made this move… which is probably one of the many reasons why Intel decided to do this as no one is really pushing x86 software on mobile devices right now…

      1. Yeah. They are really pushing WIn 10 mobile. The only thing which hasn’t arrived is the death notice.
        So to be clear you are saying all these recent universal apps announced work on the mobile phone OS utilizing ARM chips already?

        1. So long as it’s a true Universal App, not using bridge or other shortcut, and the developer doesn’t specifically block it being available on certain platforms then yes… They will run equally as well on ARM as on x86, provided of course the performance of the device is sufficient for running the app…

          Moving forward, MS basically intends to be hardware agnostic to not limit their options… It’s just most of their legacy software is still holding them back but that’s one of the reasons why they want everyone to update their apps to Universal… While also pushing for Cloud services that are of course also naturally hardware agnostic…

          MS wants to eventually change their ecosystem to the point that it’s more a question of what form factor you want to use the apps than what actual hardware it’s running on…

          Thus the introduction of Continuum to allow even the same app to change how it’s optimized for specific form factors.

          Of course, the actual execution of this and how effective it will be remains to be seen… along with how well this approach will be received in the long run.

          Anyway, Intel is on its own if they want to expand their market beyond their traditional markets… but it looks like they’ve put that on indefinite hold for now and are hoping the 2 in 1 market holds them out for now until the next big market trend shift…

          1. I think you are mis-stating it slightly. It won’t run on all platforms unless blocked. It will run on those platforms if developers enable it. And that’s not just flicking a switch.
            However enabling it takes some extra work on the part of devs. And that is not worth it for Microsoft’s efforts in mobile – which are again running on fumes.
            Your point is not without validity though.
            MS has done a lot of work to get Win 10 running on a lot of different form factors. I think their play would have been for the next gen cell phones to run on x86 and run pure Win 10. This would have been their best shot at getting app parity closer to kick-off for the next push in mobile. At solving the old chicken and egg problem everyone has when trying to start a new ecosystem.
            You are probably right though that they will simply redouble efforts at being able to run on ARM or x86 equally well. Right now I think Android is further along with that. That will make their mobile play all the more important as it’s clear that Google is coming for the laptops/desktops in the long run.

          2. Again, the issue for devs is mainly the form factor… the hardware doesn’t matter much except for certain cases like games requiring specific GPU’s to run, or a map app that requires the device to have a GPS, etc.

            So the CPU doesn’t really matter to Universal Apps, this is because the platform provides a common API, Kernel, tools, programming language, etc. for all devices running any Windows 10 based OS…

            https://az648995.vo.msecnd.net/

            This allows Universal apps to run on anything, so long as the dev enabled support for a given form factor…

            Windows 10 Mobile can already run on either x86 or ARM, for example…

            You are right that devs still require additional work but the developer tools are designed to handle most of that work automatically and so leaves the devs with primarily work they already had in making their apps unique, and optimizing the UI, etc.

            You are also right that MS won’t be getting much, if any, support for mobile apps but the thing about using a Universal Windows Platform is that apps don’t necessarily need to be for a specific device type.

            While it is also in the invested interest to support multiple form factors given that this ecosystem includes devices with some overlap… Like the XBox One has some overlap with how mobile apps are optimized and unlike MS’s mobile efforts the XBox One is a successful platform…

            Say, for example, Facebook makes a Universal App… it’s unlikely they’ll lock it to just desktop users and as long as they don’t block it then it should be available for any platform that can run it, which should be all of them except IoT…

            Mind also the 2 in 1 market, it overlaps the tablet and desktop UI, So that’s more apps that can also be used for mobile…

            So, unless the Universal Windows Platform completely fails for all form factors then it’s still possible for MS to leverage the rest of their ecosystem to give their mobile at least a fighting chance…

            Sure, a lot depends on the devs but the platform itself allows for it and that’s the point…

            While we’ve yet to see what results may come with the bridge tool to port iOS apps and the Win32 bridge has barely been available for long, or see if they ever bring back the Android bridge tool that seemed to work too well as some reported getting the Play Store working…

            To clarify, no one is saying MS doesn’t have a very steep uphill climb ahead of them but the old ARM vs x86 issue is just for their legacy support and if the Universal Apps succeed then they won’t have to worry about that going forward…

            Intel’s decision thus mainly only effects them in regards that there won’t be much progress in the cheap low cost PC tablet market and they may have to change the dividing line between mobile and desktop OS PC devices…

            Previously, with Windows 10 release, MS set the range to phone sizes and mini tablets 3″ up to 7.99″ are covered with Windows 10 Mobile and devices larger than 8″ are desktop Windows 10…

            But without ATOMs that divide has to shift to 10″ and larger instead of 8″ and larger… Though, the tablet market has already been in decline the last two years and the market was already moving away from 8″ Windows tablets…

            While MS has yet to push Windows 10 mobile towards tablets but Continuum also allows it to support laptop docks… So, in a way, you can say MS saw this coming but the market trends were suggesting it for years…

      1. ARM never would have taken off on Android if not for the early success of iPhone. So Apple basically kicked off the ARM revolution which killed Intel.

    1. Intel Killed Intel.
      The one thing no one has talked about is the fact all ARM products have a 1 OS life span unless they cost as much as a desktop. You buy one and replace it if something doesn’t work right. My Android 2.x tablet still turns on…. You can generally upgrade Intel based products for years, there is still P4 machines running companies!

    1. Not necessarily, they could just opt for the Apollo Lake Celeron/Pentium instead… Still ATOM, still updated to Goldmont architecture, etc. but still pretty cheap and low powered for good battery life…

      1. Potentially, but I doubt it. Being Braswell’s successor, Intel appears to be aiming Apollo Lake at “cloudbooks” and yoga style 2-in-1 devices. I don’t think I’ve ever seen a Braswell tablet.

        1. Mind that was partly due to the fact Cherry Trail was cheaper and lower power than Braswell and thus more ideal for tablets but Apollo Lake introduces a lot of cost cutting benefits to OEMs that Braswell couldn’t offer and it’s not like they have much choice now…

          Apollo Lake can also still be configured with mobile parts, eMMC, LP-DDR3 or LP-DDR4, etc.

          https://images.fonearena.com/bl

          https://www.techgiri.com/wp-con

          Add the improvements over Braswell and it’s at least a more viable option now…

          Besides, even Core M has made it into some tablets… Even if they’re cheap China tablets…

  17. That sucks. I love these low cost Windows tablet Chinese vendor has been pushing out.

  18. Is this the end of $100-200 netbooks again…? Currently using a 10″ Z3735F 2in1.

  19. auh man! I was realy hoping to pick up a Pipo KB1 as a toy machine to play around with… does this mean that the supply chain for new Atom based products is cut? Will manufacturers be canceling new products?

    1. The cut off is mainly for phones… anything else remains to be seen but they’re not cutting off ATOMs right away, the Cherry Trail models will continue to be sold. There just won’t be a Broxton update for them… Only Apollo Lake but that’s for slightly higher range devices.

      We can still see low cost tablets with Apollo Lake but the product range will shift more towards 2 in 1’s or low cost laptops… and anything smaller than 10″ is now unlikely…

  20. Intel is a massively dishonest company they deliberately hide the specs of what people are buying in order to pass off garbage they call Adam Celerons now and they don’t post the clock speed of any of their CPUs when you buy them in the store

  21. It was so confusing comparing cherry trail to the celeron Braswell stuff

  22. Makes sense. Intel was already calling some Atom based chips Celeron and Pentium. Now they’re fully getting rid of the Atom brand which has had a negative connotation for a while and re-branding the chips Celeron/Pentium.

    Too bad having “Atom” and Core based Celerons/Pentium is still confusing.

  23. I don’t know how to feel about this. It’s understandable Intel wants to do away with the low cost no profit Atom chips. It was unnaturally cheap, obviously it was to grab a market share. And they did. In the last two years I’ve never once considered getting an Android tablet, they all seemed inferior to a ~$100 BayTrail tablet. I did use such a low cost tablet for two years as my main to-go system, and I was very pleased with it. It did not replace my desktop, but came pretty close. And for less than $200…

    I’ve changed to a more expensive Core M system since then. I would probably never buy a $100 Atom tablet, simply because I don’t like to go backwards (thou I had my eyes on the Chuwi Hi8 Pro if nothing more than as a backup and curiosity). If my current machine breaks, I’ll only buy an even better one. So I was done with Atom myself. But there are many friends and colleagues I try to convince to replace their old Android tablets (speaking 3 year old Galaxy Tabs here) with a Windows tablet. For less than an Android tablet’s price they get the full Windows thing with FullHD IPS screens and very good processing power. Most of them are capable of dual-boot, or we can install Remix OS if we want to. Well, this will go out the window in the near future. Why should they spend – say – $400 on a Windows machine, if they were good with an Android so far? Or an iPad mini?

    Even thou Atoms were synonym of low-cost low performance machines, they were actually pretty capable! A standard 3735F is every bit as powerful as a late Core 2 Duo system with way better power efficiency. In fact most office machines don’t need more than that, only slightly more RAM, and they are good to go! For years this will be totally acceptable processing power!

    Compared to that the Celeron and Pentium names are faded. They are old, as brand names. The Pentium is 23 years old! And the Celeron was what Atoms are now: the cut-back cheap version of the big ones. In my head an Atom sounds better than a Celeron: I know what to expect.

    Speaking of what to expect, calling basically Atom chips Celeron or Pentium is nothing new. The J1900 Celeron for instance is an over-powered BayTrail Atom (10W TDP), and as such a pretty okay low-end CPU for a dirt-cheap notebook. The J1800 is basically the same thing, but with only 2 cores, and thus the processing power cut in half. I would not recommend it as a personal machine, probably good for a single use kiosk or embedded system. Still a Celeron, the 1037U is an IvyBridge CPU with about the same power as a J1900. What a total mess, right? The J1800 should be a Celeron i3, the J1900 a Celeron i5 and the 1037U probably a Pentium i3 (higher architecture, but cut back to the lowes level). But these are only names and I suppose Intel will do something about this if they drop the Atom name.

    1. 1037u is a nice, sweet little cool spot in the market. It has same passmark as j1900 but delivers it from 2 cores rather than 4. And it runs well as fanless htpc

      1. Yep, the Celeron 1037u (22nm) at 1.8Ghz is very snappy. Getting harder to find mini-pcs or itx MBs with them now, and the price is usually jacked up because people have figured out it was a classic design.

        Another great design is the Baytrail 3735F (22nm.) Perfect for a little windows tablet or 2-in-1 with an IPS screen for browsing or viewing.

        Since 22nm, maybe even 32nm, cost per transistor has been going up. Moore’s Law is dead. intel is adding more features to new CPUs, but nothing that substantially improves user experience.

      2. I’m using it on my laptop. (brand name is clevo). nice for the price.

    2. Celeron i3 has been always slow. But Intel is a world wide company and his chips always better than the other chip making company

  24. they tried this trick with chromebooks to kill netbooks once, we see who won. now they try doing the same game again. maybe they wlil have to lay off more personnel in future.

  25. Giving up on the cheap end of the tablet market doesn’t mean the expensive end will remain. What the history of the technology sector seems to tell us is that it’s far, far easier for a cheap, ubiquitous processor to erode the market space of higher end offerings than the other way around. Intel themselves should surely be very aware of this is that is precisely how derivatives of the x86 processor gradually drove high-end server CPUs into ever smaller niche areas (or killed them completely).

    The two areas where Intel continue to make decent profits are servers and the desktop/laptop market (and I think the dividing line between high-end laptop and tabled is a bit fuzzy). There are some major barriers (mostly around software) which will make it difficult for ARM architecture systems to supplant x86 in those areas, but it seems quite possible that there will be a continued encroachment as the way in which functions are delivered changes over time.

    1. I write this from a 4700MQ 17″ laptop with 3 SSD’s a dedicated 2GB GFX and 16GB RAM. There is a world of difference between a high-end laptop and a low-end one in terms of price and features.

      For Gaming and server workloads I’ll hapily concede the laptop and nettop are poorly conceived form-factors. For gamers the laptop form-factor doesn’t really make sense as it’s not hard wearing enough, and for large multi-tennant systems again, it’s power-efficient enough, but not power effective enough.

      For a coder, a systems / technical architect, someone that wants portable power and only an hour or so of life in-between charges to run some beefy apps, multi-task without lag, laptops can be incredibly powerful!

  26. > While you can’t run the full Windows 10 operating system on an ARM chip

    (Well, IMHO Windows RT which did this (and I still miss my Surface 2 RT which was/is a great tablet) was killed by the media.)
    Anyway, Windows 10 Mobile does run on ARM and why should it not be used for tablets as it is just displaying apps in a different size which Continuum already does 😉 And Windows 10 Mobile IS a full Windows OS depending what you defined as “full”. Of course it does not run x86 programs (only via Remote desktop which may be fine though) but the OS itself should basically be the same and all store apps including a now good Office runs just fine. And same as no x86 Windows program runs on ARM guess no x86 Linux program runs there either. But maybe I am wrong about that?

  27. I was wondering when they were gonna get their shit together. They have a ridiculous product labeling and marketing spread, with things like a core i3 outperforming an i7 and etcetera. They’re very likely killing the Atom because of it’s name and association with low and slow devices (evidently not at all true but perception is the rule of law) They will still play in the low cost low power pool but under a different product name scheme.
    And Broxton had to be done away with because it’s stupid to develop something that underperforms something you already have, witch is the Core M series. Rather than develop something for that price point they can just as easily repackage crippled or underclocked previous generation Core M dies and place them in the sub $200 market. This way they save a ton on development and it’s not like they’re reinventing the wheel here, Nvidia has been pulling this shiznit for decades, the current Kepler chip found in the 980, is staying in production and coming back next generation, slightly underclocked and with a few CUDA cores crippled, as the new 1050, 1060 and the 1070.

    1. Yeah, my hope is this amounts to a rebranding, if new low power Intel chips fill the gap currently occupied by Atom. Often the names are irrelevant anyway, it’s not like Pentium today means the same thing as it used to.

  28. A low power soc from AMD would have been more interesting. But AMD doesn’t have the money, stayed too long at 28nm and contra revenue killed any chances for them to get into a tablet. Let’s hope that in a year from now a low powered AMD SOC with x86 Zen and Polaris GCN will be good enough to make a little game console in the form of a 2 in 1 device with a price close to $200.

  29. Next Asus T100 will be more expensive 🙁
    cherry trail atom is good enough for most people, Intel forcing everyone to buy more expensive celeron/pentium/core-m

    1. Well, the Celeron/Pentium aren’t as expensive as the Core M’s and they’re all still cheaper than the official Core series…

      Note, most low cost Chromebooks are configured with a Braswell Celeron right now and they’re still pretty affordable… They just won’t be a cheap as mobile SoCs meant to compete with ARM but that could also be a good thing as we may see better parts used than mobile only hardware…

      Something better than a eMMC, for example, could do the Asus T series some good… in both capacity and performance… Using the same parts as a phone has its downside…

      1. Note an emmc still beats regular hard drives that many computers still use (especially on non sequential). Though I certainly wouldn’t say no if we start seeing full SSDs in future 10″ 2-in-1s.

        1. Sure, but the advantage is mainly in the short read/writes… eMMC’s are basically around where SSDs were back in the early netbook days when we only had up to SATA 2.0 options…

          Mind, it wouldn’t be as noticeable but for the fact that mobile hardware also usually means the system is configured with minimum RAM and things slow down even for SSDs when they’re forced to not run on RAM…

          SSDs combined with more RAM than mobile SoCs are usually configured with yet means there would be a lot less bottleneck to performance… versus running a mobile SoC system on eMMC with only 1-2GB of RAM usually…

          Though, some of the latest eMMCs have started to offer improved performance, and mobile SoCs are starting to finally be offered with more RAM, but it will still be years before they get around SATA 3.0 range…

          Mind, eMMC is a single chip drive that shares more with a SD Card than a SSD… eMMC even stands for embedded Multi-Media Card and started out even slower than a modern HDD.

          So a lot of the perception was from a few years ago before they finally got eMMCs up to SATA 2.0 range of performance but that bottleneck was often blamed on the SoC ATOM instead… While many cheap devices are still given the older and slower eMMC drives, which doesn’t help the perception…

          But configure a ATOM with a real SSD and at least 4GB of RAM and there’s a definite noticeable performance difference…

  30. who cares.. Core M is pretty much the same thing anyhow just a bit more pricy

    1. Uh, no… Core M is a more powerful processor, even with limited TDP to be in similar range as a ATOM doesn’t change the fact that a Core processor has 50% processor efficiency advantage over a ATOM processor clock for clock.

      While a Core M still uses more power overall (doesn’t power sip as much or go into the really low power states that mobile processors go into) and can still generate more heat and is thus why Core M systems tend to worry more about over heating and throttling despite being only dual core processors and not quad cores like most ATOM offerings…

      ATOMs are just cheaper and even more low powered for even better battery life… Core M’s can’t be put into anything smaller than a 10″ tablet because they’re barely in the mobile design threshold and can’t go any lower like a ATOM can… So you’ll never see a Core M phone or anything else that small, even the Intel PC Stick Core M version is larger than the ATOM one and requires a active fan cooler to work…

      A Core M can also unlock to higher TDP if the OEM allows it, such as with a dock that provides additional cooling and extra power to put even closer to what a Core i3 offers, which a ATOM could never do because it’s only designed as a low power chip…

      So the price difference may be more than the performance difference but there is a difference… Core M may have a similar target SDP but that’s about it…

      1. well they meet a similar market since they can both run fine with passive cooling intel seems to just be cutting their losses since the realized that they cannot compete with ARM processors and people are just snagging the cheaper tablets..now that their clearanced i bought one just to read comics on and it is quite capable for basic tasks.

        1. Yes, Intel is cutting their losses but it’s also following market trends and IMO accepting that they need significant advances before a major push into mobile can be profitable for them.

          However, there’s still a need for low cost and low powered solutions for the 2 in 1 and low cost laptop markets that the Core M doesn’t fill…

          It takes more than just being able to run on passive cooling to be in that category and Core M barely can run on passive cooling anyway… It requires larger and better engineered heat sinks than the ATOM based Celeron/Pentium SoCs, which in turn just add to the cost of the system.

          So it’s much harder to be low cost with Core M, even the Cheap China tablets have a significant price difference and it’s not all just because they charge more for the chip…

          It’s just a less significant difference now that we won’t have the even cheaper and lower powered updated mobile specific SoCs to compare to anymore, but the difference is still there…

          The Celeron/Pentium ATOMs will just become the goto product for the low end for now… They won’t be quite as cheap but they won’t be as big a price increase as the Core M’s would be and the battery life can still be good enough with the updated Goldmont architecture…

          1. looks to me like the low-end laptop market will be superseded by ARM based laptops once enough companies rebrand them those All-Winnder based laptops looked pretty neat.

          2. Hmm, maybe… If MS is successful establishing Universal Apps then a lot of desktop users will be able to make the switch to devices powered by mobile SoCs and not have to worry whether it’s x86 or ARM based…

            Mind, while Linux Desktop is already a option many ARM platforms don’t have good Linux desktop support and there aren’t any clear signs that’s going to change any time soon…

            But it will take time either way, ARM SoCs are still not quite ready for filling that role yet. Though, they are getting very close now… Maybe another two years, but by then Intel will already be on 10nm and getting ready for 7nm and if the ARM manufacturers can’t advance past 10nm then Intel could make a comeback then with a clear FAB advantage…

            Admittedly still a long shot but the technology is getting to the point that we may see a game changer that will upset the present balance eventually… allowing either ARM to explode onto the desktop market or Intel to go into markets we never thought they could before… We’ll see…

          3. I was talking about those Remix OS All-Winner based laptops… i am kinda hoping for some Mediatek based laptops as well

          4. Yup, so was I… Allwinner has been accused multiple times in the past of violating the GPL license by not providing Linux/Android kernel source code or U-Boot source, and by using LGPL-licensed code within their binary blobs, etc.

            And they’re not the only ones unfortunately… There’s a reason there hasn’t been any long term products with ARM that doesn’t just get replaced completely instead of upgraded each year or two… even in the much more financially backed Android markets, rapid end of life is all too pervasive…

    1. Both Zenfone 2 Laser (released Aug ’15) and Zenfone Go 5.0 LTE (released Apr ’16 in India) use the Snapdragon 410.

      1. ZenFone 2 (ZE551ML) is powered by a 64-bit 2.3GHz Super Quad-Core Intel® Atom™ Z3580

        1. The point, I believe he was making, is that not all ZenFones necessarily came with Intel to begin with… It’s just now pretty much assured that the next version won’t be Intel based… At least not the SoC, as they may still use other Intel parts like the Cellular Modem chip and Intel is definitely still moving forward with 5G advancements…

        2. Go to gsmarena.com and search on “asus zenfone”. All the Zenfone variants to date will be listed. If you peruse the list, you’ll see a definite cut-off date when every Zenfone released before that date will be Atom, and every one after will be non-Atom. The Zenfone 2 was released on Jan ’15, and was still in the Atom grouping.

  31. AMD should have the budged Stoney Ridge platform soon https://browser.primatelabs.com
    Sure higher power 5-15W or so but perf per $ should be interesting.

    Anyway, Apollo Lake is the new Atom, no matter how they call it. What they are doing is exiting the smartphone SoC business.

  32. I think the above Liliputing piece exaggerates the “death” of the Atom (for clicks?) Intel is cancelling the smartphone SoCs, the Broxton and the low-end SoFIA. It’s not cancelling Goldmont, the Atom core that was just recently announced, and will go into upcoming notebook/netbooks as well as tablets.

    Head over to Anandtech for a more detailed and accurate reporting piece.

    1. Well, it’s still a big deal they’re basically giving up the name “ATOM” and ceding the mobile market to ARM… But yes, Apollo Lake is still part of the ATOM architecture series, it just gets sold under the Celeron/Pentium branding and doesn’t specifically target the mobile range but is still low cost and low powered…

      Given the declining tablet market and emphasis on 2 in 1’s and low cost laptops… It seems the logical choice for now, especially as we’re still looking at about 2 years before they can move to 10nm FABs…

      1. The law of diminishing returns applies here. I think transistor size has shrunk to the point where (at least for the consumer market) we have reached the end of the line. I don’t expect we will see 10nm in the consumer or desktop computer market — ever, nor is there such a need to attempt to make chips with that size transistors. Moore’s Law is definitely dead as the trouble chip makers are having shrinking their transistors are showing, any ability to physically manufacture 7 or 10nm chips will cost so much more that those chips will be limited to mainframes and cloud servers costing millions of dollars each.

        1. If it was just a matter of transistor count then the law of diminishing returns would have taken over over a decade ago… No, we will definitely see 10nm in the consumer and desktop computer markets… and yes, there’s a need, until we get super computers on a single chip we can implant into our heads they will continue to advance the technology…

          We’re still a long way from having robots with True A.I., virtual reality that can actually fool our senses, labs we can carry with us to do DNA testing, etc. on the go, efforts to make real life tricoders still need a big breakthrough in mobile technology, among many other things that will continue to provide the demand to push the technology…

          Sure, cloud computing will offset a lot of the traditional reasons to push technology but like the ones I listed there’s still plenty of other things that won’t improve with just Cloud solutions and we’re still a long way off from the singularity…

          And no, the FAB’s may cost increasing more but the actual chips costs increasingly less! Computer products have gotten cheaper with each new FAB advance, not more expensive!

          It just has gotten more expensive to create each new FAB advance but the benefits still outweigh the costs and the consumers still benefit and that’s not going to change any time soon…

    2. The new Goldmont chips will be branded as Apollo Lake Celeron and Pentium processors, not Atom.

      Basically, they replace today’s Braswell chips.

      1. I don’t see where in the Intel statement that it’s phasing out the Atom brand, only that Broxton and SoFIA are cancelled. The cancellations aren’t a surprise, given the troubles and monetary losses Intel has incurred trying to get into the smartphones market. Regardless of whether the Atom brand lives or dies, the important point is that the underlying tech will survive on the PC side, a fact that, frankly, your piece above misrepresents.

        1. Considering Intel has yet to announce what comes after Goldmont… It’s not exactly a fact that it has a definite future to look forward to either…

          While this article did include the fact that Apollo Lake is still being released…

          So misrepresents is an exaggeration… I agree, it could have made it more clear but this is still a big game changer as the ATOM had been primarily focused on the mobile market ever since they ended the netbook market and that’s the only part of the market where it was Intel’s only offering…

          Apollo Lake has to share with Intel’s other products in the Celeron/Pentium range and it remains to be seen how well it compares to the Core M, which is going to get another update next year as well…

          So there’s a definite question mark on the future of the ATOM…

          I already expressed my doubts that Intel will forever leave the mobile market segment alone but if they do a redesign then they may actually replace the ATOM with something else… at the very least a major redesign may call for a name change…

      2. So Intel is phasing out Atom brand because it is almost synomymus with “slow”?

        1. No, if that was the case then they would just rename it but they’re actually stopping the whole market segment for Smartphones on up to small mobile tablets…

          So it’s the actual market segment that they’re dropping out of… The ATOM branding just happened to be still used for those segments but has long been phased out of the others that the ATOM architecture is also used in…

          Though, they made no mention of the server or embedded markets but those are usually custom markets and don’t reflect their usual consumer market strategies…

    3. Internal benchmarking showed that goldmont is IPC equivalent to A72. They are watering-down the celeron/pentium name when they rebrand.

      1. 1) Pentium/Celeron Branding just means the product is intended for the budget/value PC range and has little to do with performance… Products sold under this branding are either naturally low cost or are gimped/crippled down to be low cost, in either case the branding is for cost value and not about performance…

        2) What’s wrong with a A72? ARM has come a long way in performance…

        It basically has what? 20 to 60 percent increase in IPC over Cortex A57… Sure, a Core product will still easily beat that but under mobile power constraints it actually holds up pretty well in comparison.

        It helps that a Core M will be limited to duo core but the mobile SoCs can go 4, 8, 10 cores easily…

        3) What internal benchmarks? Apollo Lake isn’t due out till near end of the year and Goldmont is a scalable and customizable architecture much like ARM. Even if there were early benchmarks it could have been for some entry level device meant for something akin to a phone… We’ve yet to see any benchmarks for Apollo Lake, which also won’t be limited to mobile SDP/TDP…

        So, even if equivalent to A72, it doesn’t mean the performance will have to be lousy…

  33. Next is IOT. Set the alarm for 5 years from now. They make great laptop and server SOCs. They just need to stick with that and save money with a smaller number of employees.

  34. The “prosumer” market just isn’t big enough to meet Wall Street expectations. For devices that sell in the volumes Wall Street demands ARM is plenty good enough. It has nothing to do with what we might want. That means things like conventional PCs will become more expensive in the non-free market driven by demand for growth by the investment sector. Society will just “go gentle into that good night” as we slouch further into consumption driven neo-feudalism, the endgame of Reaganism.

    1. komrade keith would prefer a more centrally-planned/regulated PC industry that would have us all still using Osborne 1s, of course…

    2. You have it exactly backward. The PC was unnaturally cheap for years because of the Microsoft / Intel duopoly. The average user never needed Windows and the complexity it brought but there were no viable options because, other than Apple (with issues of its own like insanely great margins), preloading anything else on a PC was (illegally) forbidden, excluding the exact market segment who would have been open to a simpler option. Once Google blew that up with Android people voted with their feet in the billions, they didn’t migrate to it; they stampeded.

      Now you really need to understand that the PC and the relentless grind of Moore’s Law that gives everybody these cheap and powerful PCs depends on massive volume to generate the cash flow needed to fund the huge R&D budgets required to make that happen. Guess what my Socialist friend, when the sales dried up so did the profits that were driving everything to faster, better and cheaper. Now we get faster, better and cheaper tablets because that is where the volume sales are.

      Intel tried to buy their way into tablets and phones by selling below cost but eventually even they had to stop losing money on every sale and trying to make it up on volume because they realized there was no new monopoly at the end of the trail. Their only selling advantage vs ARM is “our chips run Windows” but Windows 10 bombed, right behind Windows 8, nobody wants em and Intel grew tired of waiting for another “winner” from Redmond. These are all market driven forces working themselves out.

      1. Not true at all, people didn’t leave PCs for Android, they bought Android phones to use alongside their PCs. And the advantage of phones is one of their portable nature, not due to Android versus Windows – because hardly anyone is running Android on a laptop or desktop. Nor did it start with Android – before that there was Symbian, or other platforms like S40, BREW etc. Saying the average person doesn’t “need” Windows is like saying they don’t need Android. Back in the 80s and 90s there were plenty of home computer platforms.

        I’m pretty sure the smartphone industry is being driven by R&D budgets at least as big as what drove PCs. PCs are still miles faster than tablets – while the sales of the latter are declining also.

        Windows 10 has been a massive success, while even Windows 8 sold 100 million in 6 months.

        1. Good points. Android succeeded mostly because it was there first and Windows wasn’t ready. Now, with Win 10 out and UWP getting ,more popular, the situation is very different.

  35. This makes no sense at all, just when they were making some decent headway into getting more market share in the tablet space with Baytrail and now Cherry Trail – they do this?

    It just got to the point were people were really starting to notice Intel powered tablets , and the myth that they would suck your battery dry had finally dissipated. People were starting to really take notice and more and more tablets with Intel SoCs were coming out.

    Now they just give up on it. It’s like they went more than half way across the river then decided to turn back. WTF

    1. Well, there are factors like the tablet market has been declining the last two years in a row… Much like the already declining PC market…

      The only real market growth has been 2 in 1’s and low cost laptops like Chromebooks… But those are actually better served by the Celeron/Pentium branded ATOMs…

      While progress hasn’t been very good for the phone market, where the bulk of the mobile market is focused on… Especially as the phablet market has eaten into the tablet market…

      Add, many believe Intel needs to get onto their 10nm and maybe even their 7nm FAB before they can really compete on even footing…

      ARM is naturally cheaper technology right now, a FAB advantage is needed to put Intel on equal costs with ARM products and not have to rely so heavily on subsidizing their products, which is costing them a lot… While a x86 advantage doesn’t help if people won’t run any software that’s exclusive to x86 of the majority of devices that ARM competes on…

      Sure, they can continue to burn money and compete that way but I think they’d rather compete when they’re in a better position than they are now…

      The 14nm FAB will continue on for another year before products based on the 10nm FAB even starts to be introduced to market and if there were any indication that the Broxton update wasn’t going to be as good as they hoped then it does make some sense for them to cut costs and wait for a better time to really push again into the mobile market…

      The costs to advance to the next two FAB advances will also be quite huge by themselves and with declining markets to deal with that money is best reserved for those FAB advances…

      Think of it more like building a bridge across a river… just because they may have a design flaw that forces them to destroy part of the construction doesn’t mean they’re necessarily going to never build the bridge ever… but a redesign may be in order before progress can begin again…

  36. Next year’s Intel still has to beat the competition. The competition is last year’s Intel. That is what the demand demands.

  37. I hope this isn’t going to be death of netbooks all over again – where there’s no choice again for low power ultra portable PCs, and the only choice is fisher price ARM devices again. Microsoft will have to bring out an ARM based Windows again… (and as we know people would rather have x86).

    I don’t mind paying more for higher performance, but I do like the ultra portability of 10″ / 500g devices, and it seems all of the Intel core devices have always been larger. Will new low power celeron/Pentiums be a viable successor, or will they perform poorer on battery life?

    There might not be much profit in low cost devices, but there’s even less profit if ARM takes over. Maybe they’re right to get out because tablet sales are plummeting, but does this mean they’re also giving up on mobile phones?

    1. Well… It’s hard to say what Intel’s long term plans may be at this point… Like you I doubt they will just give up on the entire range and cede forever to ARM but it’s not like they are gaining much market share and with the cost of subsidizing is costing them quite a lot and it may not be worth it at this point if they can’t offer a game changer any time soon…

      Many have longed reasoned that Intel needs to push into their next two FAB advances before any others can catch up before they can really compete with ARM and that may be what they’ve decided to do…

      The 14nm FAB will go on for at least another year before we see the first of the 10nm FAB and the 7nm is at least another two years after that… So this may be a 2 to 3 year hiatus…

      Mind, the cost for advancing a FAB increases each time and with the declining market growth it’s increasingly hard to justify those costs…

      Add, Intel is still playing catch up with GPU performance in the mobile range and Broxton may not be the big update they were hoping for as well could have also given them reason to take this choice of action as another incremental update like Cherry Trail may have hurt them worse than just taking a hiatus for now.

      While focusing their efforts on the Celeron/Pentium range allows them to not only lower costs significantly but also put more resources into R&D to speed up advancing to the next stage…

      It helps that this doesn’t mean they’re giving up on low cost ATOMs, they just won’t be focusing on the mobile range or products that are necessarily branded as ATOMs… Though, there may now be significant overlap between the Core M and ATOM development that may or may not lead to a eventual merger at some point…

      Mind that starting with Bay Trail that the ATOM started to adopt some Core series technology like the GPU, more so with Cherry Trail and Broxton would have done it again…

      Mind, also that with the declining tablet market it makes more sense for them to focus on the 2 in 1 and low cost laptop markets that are the only growth market segments right now…

      So, if a merger… at least on the low end… is their end game then this action does make some sense for now…

      As it was expected that Intel would eventually offer a fully scalable solution… Broxton was suppose to be the first introduction of a scalable and customizable architecture but if they decided to skip the introduction and wait for the final version then it could mean the next time we see a mobile range option from them it will be for a chip that can scale all the way up to desktop range without needing to switch between processors or at least one that can easily integrate to more powerful ones by just plugging the device into a dock… along with 7nm FAB could mean a real game changer, just not any time soon…

    2. “Microsoft will have to bring out an ARM based Windows again… (and as we know people would rather have x86).”

      They already have one — Windows 10 Mobile. They’d argue that with Continuum you have a full desktop OS running on ARM, so in a laptop it’d essentially be like a permanently docked phone. Of course you lose access to the vast world of x86 software, but (unlike RT) you will at least have access to a large percentage of current and future Windows software.

      I’d agree that regular old Windows 10 on an X86-compatible chip would be vastly better, but a Windows 10 Mobile laptop should be a lot less gimped than RT was.

  38. I’m surprised my surface 3 runs so good but I realize it only has 2gb of ram and my phone has 3gb of ram and came out in 2014…

    1. Well, you could have gotten the version with 4GB of RAM, 2GB was just the minimum configuration option, but later, once they start the switch to DDR4 RAM, they should start offering 8-16GB, especially if they use a Apollo Lake in the Surface 4… along with a real SSD instead of eMMC…

      Mind, Broxton would have only improved eMMC support to 5.0 but the Celeron/Pentium branded ATOMs aren’t limited to just eMMC drives… While still being cheaper than Core M…

    1. Still coming with Apollo Lake… They’ll just be for the Celeron/Pentium branded markets but those still cover 2 in 1’s and low cost laptops…

      We’re just unlikely to see any dirt cheap Goldmont products, like sub $150 tablets, etc. as the Celeron/Pentium range isn’t subsidized like the mobile SoC range… at least not as much…

      And they’re pretty much exclusive to the Celeron range for Chromebooks already… but maybe we’ll see some Cloudbooks updated to Apollo Lake sometime next year…

Comments are closed.