I reckon the VP9 HW decode is the biggest feature most people would notice on this release. Youtube acceleration will help this class of device fairly significantly.
People who primarily use computers to watch stuff and primarily use their heads for balance and to wear hats will rejoice.
I don't wear hats and can't help but wonder if the platform is not too limited in terms of IO. PCIe lanes are few, and bandwidth is low too, since it is ancient v2.0 (its from 2007!), with v4 incipient it is kind of odd to release a brand new platform stuck at 2.0.
I guess it is a measure to keep power usage low, since v3 switches to the more taxing 128b/130b encoding. Still, 3 GB/s theoretical outta be enough for fast storage and the generic platform peripherals, but discrete GPU seem out of the question.
Without Intel's Atom and their "contra revenues", we would still be looking at $500+ x86-powered tablets. Examples http://www.anandtech.com/show/6522/the-clover-trai... (which uses Intel Atom) and http://liliputing.com/2011/12/bungbungame-launches... (which uses AMD Mullins). By bringing the prices of x86-powered tablets down (to very affordable levels), today even more powerful Intel Core-M tablets can be bought for $300+ or below.
mind telling me where to get a new core M tablet when a single core M chip itself cost $280? The cheapest new ones I could find were around that $500 mark.
Yeah I have a Cube i7 Book a skylake Core-M 6y30 4GB RAM 64SSD (m2). I got that for about $320 USD. I've seen the price of the CPU and yes... I know right! what's going on? Nevertheless I'm so glad I got this over an atom based device! I mean to have a REAL BIOS that I can boot onto SD/USB oh what luxury!
Nooo, it doesn't cost much more or maybe even less to manufacture the lower performing chips. Their purpose is to allow them to jack up the price for a better performing chip/computer. I remember years ago, I forget which Intel chip, but they actually disabled a portion of the chip to sell a lower performing chip and the full functional chip at a higher price. I am watching the shopping channel right now and it's disgraceful how they are selling a Pentium quad 4 Appolo Lake all in one computer as the greatest top of the line Intel to people who don't know ant better. It's just an improved Atom chip that belongs in a small tablet at best
Sounds like they are using faster memory in the benchmarks. This is fine, since the platform no doubt could use every advantage it could get. My only concern would be that OEMs will skip LPDDR4 on the budget devices these will go in.
Overall, it seems the original Atom concept has lost its way, as these chips have really started competing with Core-M in terms of TDP. I wonder if there would be a strong business case to dump this platform entirely, and sell a low-margin Core-M Pentium line instead? Will Atom make a comeback when Intel finally gets its next process node going?
It's actually not a bad thing to see Apollo Lake staying outside of tablets. x86-based Windows tablets are few and far between anyway and many of the budget models suffer from a lack of RAM and storage space that made them frustrating to consumers rather than helpful. When Android devices and iPads can make do with less memory and less hassle, the point of a Windows tablet is moot in the eyes of a consumer. Among the tech savvy, the advantage Windows used to have in the past was its respect to user privacy was done away with by Microsoft during the transition from 7 to 10 (disregarding the backporting of more heavy-handed telemetry mechanisms to 7, but that's a whole other discussion).
Sure, but I was still looking forward to a non-pro Surface update. I like the idea, just can't quite convince myself it's worth what they charge for the Core-M. An updated Atom with a decent amount of memory and storage is a perfectly capable W10 tablet for $200 less than Core-M.
Exactly one year ago Intel released Pentium 4405Y, the first 2C/4T 6W Pentium part with recommended customer price of $161. http://ark.intel.com/products/89612 Unfortunately it didn't gain any traction, because is one year period it couldn't get into any actual device, which may be a reason for canceling the tablet atom parts and to end competition between two Intel's products.
Of course but the Atoms are still far cheaper and lets them improve other aspects of the device or drop the price further. For example there's no tablet in the $500 range that is built as well as the Surface 3 or has as good of a display (see the display tests here on AT). Also that's an outdated 6W part - 6W is a bit high for a tablet IMO. There's other lower-power Core chips that are way better but they also cost even more.
surface 3 says hello. It was quite a good x86 tablet. If MS updated the eMMC to a faster, newer standard, or even used UFS, it would be a great tablet. It wont happen without a SoC to put in it though, and intel was the only option.
I didn't think the non-Pro Surfaces were very good sellers. I know there was a pretty big price cut, but that may have been to clear old stock and not an indication of poor sales. The Pro models were popular for a while, but some of the corporate types to which they initially appealed seem to be backing off on the practicality of even the flagship model. Maybe that has more to do with buggy Skylake, but around my office they've lost their luster and are being swapped out for Dell laptops in what the IT guys are calling a "frustration saver" upgrade. Which isn't an upgrade at all given the specs of the replacement hardware, but a lot of people are eager to get a laptop back.
The only Surface based on the Atom core (Surface 3) was a flop. The rest of the non-RT Surface's use low voltage of Intel's big chips, which Intel just launched a new version of a few days ago.
the rest of the non-RT surfaces are the significantly more expensive surface pros, which are a different, larger size, serve a different purpose (laptop replacement VS media consumption tablet) and dont have a LTE option like the surface 3 does.
The Surface 3 wasn't that much smaller or cheaper, and didn't serve that different a purpose, which is probably the reason it flopped, was cancelled, and not replaced. The lack of Intel offering a new SoC doesn't really have any impact on that decision.
You're smoking something if you think the lack of a new SoC didn't have any impact on that. The Surface 4 needed Willow Trail to be viable. Willow Trail's tablet variants were canned. I suppose they could redesign it around Core but that would be more expensive and they've already got the Pro series.
"Tech savy" user doesn`t care about muh privacy either, he`s faceberging with both hands, eagerly telling (((Zuckerberg))) all he can, and tagging his friends without their consent for good measure.
Some of them are yes. There's certainly something to be said about throwing the entire uninteresting mess of your life to the corporate wolves so they can pick the best butt rash treatment to toss at you in a targeted advert. Some of us still try a little, but maybe at this point there's no benefit to making the effort.
Telemetry stuff's been in every OS for ages now, and (if done right) isn't personally identifiable at least. I don't see how that's a reason to go with a less capable OS (that has more telemetry if it's Android from anyone but Google).
So you think that the budget models aren't good enough (truth is there are some good models out there too) so your solution is to deny them an option for a new, better SoC. Thus killing off new affordable models - yeah that's ONE way to fix things. If you had used Windows 10 on a tablet you would know that it really isn't a memory hog in typical tablet workloads, and most of them include SD card options, though models with lots of internal storage are also out there. Either way as icrf said now we won't get a new Surface non-pro since they killed off Willow Trail.
The privacy rant is funny given that you recommend Android devices as an alternative. Do you have any idea what percentage of Google's income comes from ad revenue? 90%. Which they generate by harvesting as much data as possible, and selling it wholesale as well as serving you with targeted ads. But Microsoft is the easy target.
It's almost clear that Intel is currently on the push in making Core *the* chip, scaling from very low power devices to servers. The only part Core haven't cracked is the entry/cheap/affordable segments. I believe this is not for the fear of watering down Core's premium status. In fact I believe they already have the proper recipe for positioning Core on these cheap devices: Core m and the essentially void press coverage of anything Atom/Apollo Lake/Goldmont to wipe its existence from our memories.
Now assuming the above claim is true (disclaimer: all pure speculation), I lay down reasons why Intel still cannot do it:
What I see is that Intel can't yet make Core's yields/profitability in par with Atom's. Atom* was only a money loser solely because of Intel's contrarevenue program which I believe has long ended but with Atom continuing to stay strong in the entry level.
There's also the question of whether to cut down from Core's "base design" (if there's such a thing) to decrease die size and hence increase yields. In this respect I believe Goldmont is really a cut-down version of Core maintaining the same yield-level and profitability as previous Atoms. But perhaps it still hasn't come to the point where perf (or more appropriately, experience) could be justified as "Core-class" without some serious media backlash.
A cut-down Core could also question Core's scalability claim, essentially making it an unwanted child across the lineup. Perhaps this is where Intel's stream of new application-specific features (AVX-512, TSX, encrypted memory) can help the low end by creating more possibilities of market differentiation which in turn can be a reason to decrease die sizes by omitting them in these low end designs (not just binning).
* By "Atom" I mean Atom-class (*mont) inluding Pentium and Celeron.
I'd say this is a drastic misinterpretation of the world.
Intel bet EVERYTHING on the theory that the world cared about (and would pay for) x86. Sadly for Intel this is not true. It's OBVIOUSLY not true in phones and tablets, and it grows less true for everything else from dedicated boxes (like NAS or phone gateways) to hobby boards to even servers every day as the ARM ecosystem fills in the various remaining holes.
It didn't have to be that way --- Intel COULD have offered up Atom as a different ISA that wouldn't compete with x86, or it could have used Atom as a way to evolve the x86 towards something easier to keep improving. (Provide it as the second Intel ISA, let the ecosystem grow, then let everyone know that it's going to become the preferred Intel ISA.) But Intel was so in love with x86 that they simply couldn't believe the rest of the world either didn't care or actively loathed it.
So where is Intel now? Basically screwed. Sure they can keep bleeding money out of the high-end for a few more years, but by 2020 or so those ARM server cores will have moved from being a low-performing curiosity to a credible threat. At the mid-range, I expect Apple, before 2020, to switch to their own (better performing) SoC for Macs, at which point it becomes clear that the only reason to stick with x86 is if you really need Windows. At the low-end, well they've already lost that, and Apollo Lake ain't gonna win it back.
And so we see Intel going begging (that's not the way they spun it, but make no mistake, that's what it was) to ARM to please validate them as a foundry. If they can't make money selling x86, can they at least make money selling 14nm FinFETs? We'll see how that turns out, but my guess is, not well in terms of profitability. Around ten years ago Intel chose *x86* as its defining characteristic, not *kickass foundry*. They're trying to pivot now, but ten years ago they could have charged premium for their foundry services. Today they're basically in the same boat as Samsung and TSMC --- better at metal density, worse at packaging. I ALSO suspect they haven't yet learned the humility that's required of being a johnny-come-lately in a service business, so I suspect their are going to be a few less than pleasant customer interactions along the way, and more than a few whispers between high level execs about "well I can't say anything, but let me just point out we used them for our last chip and the next one we're going back to TSMC --- get my point?"
In Intel's defence though, every time they have tried something other than x86, they have failed. Remember i860 and the Itanium? Heck they even had to rely on AMD to to bring x86 kicking and screaming into the 64-bit world. And AMD, trying to maintain as much compatibility with x86 as possible, made as little change as they could get away with in x86-64. The transition to 64-bit was the time to clean up the ISA (similar to what ARM did with AArch64).
If Intel had shown leadership, cleaned up the ISA instead of trying to stretch NetBurst to 10 Ghz and writing magic compilers that would make Itanium competitive, the world would be in a very different place right now. And so would Intel's shareholders.
Well. that's why my suggestion (and this is not backward looking, it was the suggestion I made when they first shipped Atom) is that Intel should have defined itself as "the kickass foundry company" not as "the x86 company".
They pivoted once before, from memory to logic. But that was back in the days when they were run by engineers, not MBAs...
So basically confirmed there will be no surface 4 then, unless MS brings windows RT back from the grave. thats sad. Not everybody can afford the surface pro line
firstly, yes it would. The pentium has a 6 watt TDP, with a SDP of4 watt. The surface 3's atom is a 2 watt chip. Doubling the TDP is not something many devices can handle without a massive overhaul on the design.
The fact that the pentium also costs 5x as much is also a major issue. A jaguar is not a suitable replacement for a sentra, as it were.
Still, this is a chip class that we see in NAS units. I'm sure that adding an additional controller would be pretty trivial. May help explain why we see the 10W TDP... Intel might have looked at their volumes and found that the majority of this class of chip was shipping to devices with plugs (like NAS units) where the value of 2-3W power savings is low.
The NAS manufactures use a wide range of processors, if you look at Synology for example, at the high-end they use Xeon E3's at the low-end they use ARM processors. I have an older version, but the Synology DS216+ uses a N3060 (Braswell).
That's because AMD doesn't have anything comparable. If they can scale their Zen architecture down into that space somehow, then you'd see the tray price drop.
Just doing some math with some of Zens raw pre-production specs: 4c/8t = 66w, cut the cores down to 2c/4t = 33w, cut raw clockspeed down to half (remember, power usage isn't linear so you could wind up significantly lower) and you'd get down into the 16w range at least. Go with a low base clock of 1.5ghz and you could be in that 10w range or less... BUT we have to wait and see for a few more months
As an owner of an embedded Braswell N3700 system, I think Intel should just give up in this space. Braswell was a massive failure; I swear my cheap-o Bay Trail Lenovo Yoga 2 w/ Windows tablet is just as fast and capable in terms of everyday use.
On an AsRock N3700-ITX passively cooled board, the N3700 constantly throttles and can't maintain even passable GPU clocks - I think mine averages like 450 MHz when under stress. It's advertised as capable of 4K@30fps, which is laughable, when it struggles to even output 1080p streaming content without judder or frame skips every so often.
I should know - I tried hooking this up to my 4K LG OLED and the experience even navigating the mouse around the desktop was painful.
For a 6W chip in an actively cooled box, you would think that it would not need to throttle, but nay-nay. It's not like I'm asking it to do much, just stream some 1080p and maybe stream the occasional Steam In-home Network game.
This is the worst Intel chip I have ever bought, and I regret my purchase everyday.
I don't think goldmont was a failure, the real thing is intel not evolving the performance of atom designs on purpose so they can maintain those ridiculous high prices on simple dual core core m/i5-i7u models.
AMD Puma was a quad core with a $50 price tag, Cheetah was supposed to be closer to Phenom II/Athlon II in IPC.
So Intel blocks Nvidia from releasing an x86 chip, and then doesn't even bother having their own tablet chip? Well, okay, they do with "Core-M", but obviously those cost as much as cheap tablets all on their own.
I still think they made a mistake cutting out Nvidia, given more x86 is a good thing for Intel IMO, even if it's not an Intel chip.
And I'd much rather have Tegra Xx running real Windows than (eeeeew) Android.
"enabling Intel to reach cheaper markets and better compete with non-traditional devices based around ARM SoCs"
Come on, Ryan, you don't have to engage in "Opinions differ on shape of the world" journalism. This thing is a PoS and you know it, just like your readers know it. The ONLY place it matters is in circumstances where x86 compatibility matters and performance does not --- maybe signage, industrial control, and scammy low-end Wintel PCs?
You really think this sad piece of junk, with performance at, what, half or less of an A72 and 3x the price, can compete against a decent Mediatek or Rockchip SoC, something like a Helio X20? Good luck with that.
I would love to see how these perform against Cortex A72 and A73. My guess, not very favourably.
And I can guarantee that you can buy the fastest Cortex A72 from MediaTek and Rockchip for a lot less than these. As in, one sixth of the price of the cheapest Celeron here.
Sure, they are fabbed on 28nm, but they clock up to 2.4Ghz, so who cares. If a 28nm product can consistently beat a product on Intel's 14nm, what does that say about Atom's architecture?
x86 support you say? Well, how many of these units are going to be used for gaming? How many of them are going to be used for heavy applications like Photoshop? Not many. For everything else, general web browsing, office documents, media playback etc, Linux and ChromeOS will do just fine. And they work just as well on ARM (maybe even better seeing as ChromeOS on ARM will have better Android app compatibility).
Embedded systems pretty much all run Linux anyway. So do things like NASs. Routers started moving from MIPS to ARM a few years ago, a lot more of these devices will make the switch from x86 to ARM soon. And in fact looking at the Chromebooks that are going to launch in the second half of this year, that is exactly what is happening.
Not a pretty picture for Intel. Oh well, this will be Atom's last hurrah before being completely cancelled. At least Intel now has a license to fab ARM chips (hello XScale!)
Does this mean it's now absolutely official that we won't see a proper successor to the Surface 3? I was really hoping for one. Seeing how the SoC in the Surface 3 costs $37 and draws 2 W (SDP), I can't see how it'd be possible to make a Surface 4. These new Atoms cost $107 and draw 4 W (SDP); the new Y-series cost $281 and draw about 2.33 W (SDP) at best if I'm not mistaken. The m3-7Y30 might do the job in TDP-down with a $250 price premium. That'd be tough to swallow even if they make a Surface 4. It'd be better than nothing, I guess...
Compared to Atom (22nm Silvermont and die shrink 14nm Airmount), Goldmont is a bit of a beast. Both Atom and AMD’s Bobat/Jaguar/Puma core can fetch and issue up to two instructions. Goldmont can do three. Like AMD’s (or VIA’s) Goldmont has a full out of order execution engine. „Much of the changes with Goldmont were about improving the out-of-order execution compared to Silvermont, with a wider decoder, better branch prediction, and a larger out-of-order execution window. Goldmont can perform one load and one store per cycle, and it can execute up to three simple integer ALU operations per cycle. There’s new instruction support for hashing with SHA1 and SHA256, and there’s new support for the RDSEED instruction.“
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
71 Comments
Back to Article
A5 - Thursday, September 1, 2016 - link
I reckon the VP9 HW decode is the biggest feature most people would notice on this release. Youtube acceleration will help this class of device fairly significantly.ddriver - Thursday, September 1, 2016 - link
People who primarily use computers to watch stuff and primarily use their heads for balance and to wear hats will rejoice.I don't wear hats and can't help but wonder if the platform is not too limited in terms of IO. PCIe lanes are few, and bandwidth is low too, since it is ancient v2.0 (its from 2007!), with v4 incipient it is kind of odd to release a brand new platform stuck at 2.0.
ddriver - Thursday, September 1, 2016 - link
I guess it is a measure to keep power usage low, since v3 switches to the more taxing 128b/130b encoding. Still, 3 GB/s theoretical outta be enough for fast storage and the generic platform peripherals, but discrete GPU seem out of the question.Morawka - Thursday, September 1, 2016 - link
yeah these cores were destined for embedded anyways. NAS, DAS, etc.. those markets will be interested in these the most.ddriver - Thursday, September 1, 2016 - link
I don't think so, the features clearly indicate the intended use is consumer applications.fallaha56 - Thursday, September 1, 2016 - link
except that if it's like skylake it's not real accelerationlilmoe - Thursday, September 1, 2016 - link
Atom, as Intel positions it, needs to die. Period. Cheating consumers using from market position abuse also needs to die.lilmoe - Thursday, September 1, 2016 - link
***Cheating consumers and market position abuse also needs to die...BlueBlazer - Friday, September 2, 2016 - link
Without Intel's Atom and their "contra revenues", we would still be looking at $500+ x86-powered tablets. Examples http://www.anandtech.com/show/6522/the-clover-trai... (which uses Intel Atom) and http://liliputing.com/2011/12/bungbungame-launches... (which uses AMD Mullins). By bringing the prices of x86-powered tablets down (to very affordable levels), today even more powerful Intel Core-M tablets can be bought for $300+ or below.TheinsanegamerN - Sunday, September 4, 2016 - link
mind telling me where to get a new core M tablet when a single core M chip itself cost $280? The cheapest new ones I could find were around that $500 mark.Transkrypted - Sunday, September 4, 2016 - link
Yeah I have a Cube i7 Book a skylake Core-M 6y30 4GB RAM 64SSD (m2). I got that for about $320 USD.I've seen the price of the CPU and yes... I know right! what's going on?
Nevertheless I'm so glad I got this over an atom based device!
I mean to have a REAL BIOS that I can boot onto SD/USB oh what luxury!
AntB - Saturday, December 2, 2017 - link
Nooo, it doesn't cost much more or maybe even less to manufacture the lower performing chips. Their purpose is to allow them to jack up the price for a better performing chip/computer. I remember years ago, I forget which Intel chip, but they actually disabled a portion of the chip to sell a lower performing chip and the full functional chip at a higher price. I am watching the shopping channel right now and it's disgraceful how they are selling a Pentium quad 4 Appolo Lake all in one computer as the greatest top of the line Intel to people who don't know ant better. It's just an improved Atom chip that belongs in a small tablet at bestMonkeyPaw - Thursday, September 1, 2016 - link
Sounds like they are using faster memory in the benchmarks. This is fine, since the platform no doubt could use every advantage it could get. My only concern would be that OEMs will skip LPDDR4 on the budget devices these will go in.Overall, it seems the original Atom concept has lost its way, as these chips have really started competing with Core-M in terms of TDP. I wonder if there would be a strong business case to dump this platform entirely, and sell a low-margin Core-M Pentium line instead? Will Atom make a comeback when Intel finally gets its next process node going?
BrokenCrayons - Thursday, September 1, 2016 - link
It's actually not a bad thing to see Apollo Lake staying outside of tablets. x86-based Windows tablets are few and far between anyway and many of the budget models suffer from a lack of RAM and storage space that made them frustrating to consumers rather than helpful. When Android devices and iPads can make do with less memory and less hassle, the point of a Windows tablet is moot in the eyes of a consumer. Among the tech savvy, the advantage Windows used to have in the past was its respect to user privacy was done away with by Microsoft during the transition from 7 to 10 (disregarding the backporting of more heavy-handed telemetry mechanisms to 7, but that's a whole other discussion).icrf - Thursday, September 1, 2016 - link
Sure, but I was still looking forward to a non-pro Surface update. I like the idea, just can't quite convince myself it's worth what they charge for the Core-M. An updated Atom with a decent amount of memory and storage is a perfectly capable W10 tablet for $200 less than Core-M.vithrell - Thursday, September 1, 2016 - link
Exactly one year ago Intel released Pentium 4405Y, the first 2C/4T 6W Pentium part with recommended customer price of $161. http://ark.intel.com/products/89612 Unfortunately it didn't gain any traction, because is one year period it couldn't get into any actual device, which may be a reason for canceling the tablet atom parts and to end competition between two Intel's products.Alexvrb - Thursday, September 1, 2016 - link
If it wasn't outrageously expensive compared to the Atoms, it would have found a lot more homes.MonkeyPaw - Thursday, September 1, 2016 - link
I don't think it would ever end up costing that much for OEMs. The Lenovo Yoga 710 has that chip, and it is well spec'd and often sells around $400.Alexvrb - Friday, September 2, 2016 - link
Of course but the Atoms are still far cheaper and lets them improve other aspects of the device or drop the price further. For example there's no tablet in the $500 range that is built as well as the Surface 3 or has as good of a display (see the display tests here on AT). Also that's an outdated 6W part - 6W is a bit high for a tablet IMO. There's other lower-power Core chips that are way better but they also cost even more.TheinsanegamerN - Thursday, September 1, 2016 - link
surface 3 says hello. It was quite a good x86 tablet. If MS updated the eMMC to a faster, newer standard, or even used UFS, it would be a great tablet. It wont happen without a SoC to put in it though, and intel was the only option.BrokenCrayons - Thursday, September 1, 2016 - link
I didn't think the non-Pro Surfaces were very good sellers. I know there was a pretty big price cut, but that may have been to clear old stock and not an indication of poor sales. The Pro models were popular for a while, but some of the corporate types to which they initially appealed seem to be backing off on the practicality of even the flagship model. Maybe that has more to do with buggy Skylake, but around my office they've lost their luster and are being swapped out for Dell laptops in what the IT guys are calling a "frustration saver" upgrade. Which isn't an upgrade at all given the specs of the replacement hardware, but a lot of people are eager to get a laptop back.Cygni - Thursday, September 1, 2016 - link
The only Surface based on the Atom core (Surface 3) was a flop. The rest of the non-RT Surface's use low voltage of Intel's big chips, which Intel just launched a new version of a few days ago.TheinsanegamerN - Thursday, September 1, 2016 - link
the rest of the non-RT surfaces are the significantly more expensive surface pros, which are a different, larger size, serve a different purpose (laptop replacement VS media consumption tablet) and dont have a LTE option like the surface 3 does.Cygni - Friday, September 2, 2016 - link
The Surface 3 wasn't that much smaller or cheaper, and didn't serve that different a purpose, which is probably the reason it flopped, was cancelled, and not replaced. The lack of Intel offering a new SoC doesn't really have any impact on that decision.Alexvrb - Friday, September 2, 2016 - link
You're smoking something if you think the lack of a new SoC didn't have any impact on that. The Surface 4 needed Willow Trail to be viable. Willow Trail's tablet variants were canned. I suppose they could redesign it around Core but that would be more expensive and they've already got the Pro series.TheinsanegamerN - Sunday, September 4, 2016 - link
eh? the surface 3 started at $499. The surface pro 4 started at $749.$250 is a pretty big difference.
TheinsanegamerN - Sunday, September 4, 2016 - link
*849. $350 difference. Definitely much cheaper to get a non pro surface.Wolfpup - Thursday, September 1, 2016 - link
Nvidia's Tegra line with Denver CPUs would be perfect for lower cost Windows tablets...if Intel didn't somehow block them from using x86 :-/Michael Bay - Thursday, September 1, 2016 - link
"Tech savy" user doesn`t care about muh privacy either, he`s faceberging with both hands, eagerly telling (((Zuckerberg))) all he can, and tagging his friends without their consent for good measure.BrokenCrayons - Thursday, September 1, 2016 - link
Some of them are yes. There's certainly something to be said about throwing the entire uninteresting mess of your life to the corporate wolves so they can pick the best butt rash treatment to toss at you in a targeted advert. Some of us still try a little, but maybe at this point there's no benefit to making the effort.Wolfpup - Thursday, September 1, 2016 - link
Telemetry stuff's been in every OS for ages now, and (if done right) isn't personally identifiable at least. I don't see how that's a reason to go with a less capable OS (that has more telemetry if it's Android from anyone but Google).Oxford Guy - Friday, September 2, 2016 - link
"Telemetry stuff's been in every OS for ages now, and (if done right) isn't personally identifiable at least."You must be joking. Spying is all about personally-identifying information.
extide - Saturday, September 3, 2016 - link
You missed the point, anonymous telemetry has existed for a loooong time, but that's not spyware.Alexvrb - Thursday, September 1, 2016 - link
So you think that the budget models aren't good enough (truth is there are some good models out there too) so your solution is to deny them an option for a new, better SoC. Thus killing off new affordable models - yeah that's ONE way to fix things. If you had used Windows 10 on a tablet you would know that it really isn't a memory hog in typical tablet workloads, and most of them include SD card options, though models with lots of internal storage are also out there. Either way as icrf said now we won't get a new Surface non-pro since they killed off Willow Trail.The privacy rant is funny given that you recommend Android devices as an alternative. Do you have any idea what percentage of Google's income comes from ad revenue? 90%. Which they generate by harvesting as much data as possible, and selling it wholesale as well as serving you with targeted ads. But Microsoft is the easy target.
MrSpadge - Friday, September 2, 2016 - link
Whatever, it's Microsofts fault!;)
Oxford Guy - Friday, September 2, 2016 - link
Ads or not, we know Microsoft gladly does anything the NSA asks.OEMG - Thursday, September 1, 2016 - link
It's almost clear that Intel is currently on the push in making Core *the* chip, scaling from very low power devices to servers. The only part Core haven't cracked is the entry/cheap/affordable segments. I believe this is not for the fear of watering down Core's premium status. In fact I believe they already have the proper recipe for positioning Core on these cheap devices: Core m and the essentially void press coverage of anything Atom/Apollo Lake/Goldmont to wipe its existence from our memories.Now assuming the above claim is true (disclaimer: all pure speculation), I lay down reasons why Intel still cannot do it:
What I see is that Intel can't yet make Core's yields/profitability in par with Atom's. Atom* was only a money loser solely because of Intel's contrarevenue program which I believe has long ended but with Atom continuing to stay strong in the entry level.
There's also the question of whether to cut down from Core's "base design" (if there's such a thing) to decrease die size and hence increase yields. In this respect I believe Goldmont is really a cut-down version of Core maintaining the same yield-level and profitability as previous Atoms. But perhaps it still hasn't come to the point where perf (or more appropriately, experience) could be justified as "Core-class" without some serious media backlash.
A cut-down Core could also question Core's scalability claim, essentially making it an unwanted child across the lineup. Perhaps this is where Intel's stream of new application-specific features (AVX-512, TSX, encrypted memory) can help the low end by creating more possibilities of market differentiation which in turn can be a reason to decrease die sizes by omitting them in these low end designs (not just binning).
* By "Atom" I mean Atom-class (*mont) inluding Pentium and Celeron.
Michael Bay - Thursday, September 1, 2016 - link
Original N270 had a "really tiny and cheap die" as one of the main selling points, true.name99 - Thursday, September 1, 2016 - link
I'd say this is a drastic misinterpretation of the world.Intel bet EVERYTHING on the theory that the world cared about (and would pay for) x86. Sadly for Intel this is not true. It's OBVIOUSLY not true in phones and tablets, and it grows less true for everything else from dedicated boxes (like NAS or phone gateways) to hobby boards to even servers every day as the ARM ecosystem fills in the various remaining holes.
It didn't have to be that way --- Intel COULD have offered up Atom as a different ISA that wouldn't compete with x86, or it could have used Atom as a way to evolve the x86 towards something easier to keep improving. (Provide it as the second Intel ISA, let the ecosystem grow, then let everyone know that it's going to become the preferred Intel ISA.) But Intel was so in love with x86 that they simply couldn't believe the rest of the world either didn't care or actively loathed it.
So where is Intel now? Basically screwed.
Sure they can keep bleeding money out of the high-end for a few more years, but by 2020 or so those ARM server cores will have moved from being a low-performing curiosity to a credible threat.
At the mid-range, I expect Apple, before 2020, to switch to their own (better performing) SoC for Macs, at which point it becomes clear that the only reason to stick with x86 is if you really need Windows.
At the low-end, well they've already lost that, and Apollo Lake ain't gonna win it back.
And so we see Intel going begging (that's not the way they spun it, but make no mistake, that's what it was) to ARM to please validate them as a foundry. If they can't make money selling x86, can they at least make money selling 14nm FinFETs?
We'll see how that turns out, but my guess is, not well in terms of profitability. Around ten years ago Intel chose *x86* as its defining characteristic, not *kickass foundry*. They're trying to pivot now, but ten years ago they could have charged premium for their foundry services. Today they're basically in the same boat as Samsung and TSMC --- better at metal density, worse at packaging. I ALSO suspect they haven't yet learned the humility that's required of being a johnny-come-lately in a service business, so I suspect their are going to be a few less than pleasant customer interactions along the way, and more than a few whispers between high level execs about "well I can't say anything, but let me just point out we used them for our last chip and the next one we're going back to TSMC --- get my point?"
aryonoco - Thursday, September 1, 2016 - link
Very valid points.In Intel's defence though, every time they have tried something other than x86, they have failed. Remember i860 and the Itanium? Heck they even had to rely on AMD to to bring x86 kicking and screaming into the 64-bit world. And AMD, trying to maintain as much compatibility with x86 as possible, made as little change as they could get away with in x86-64. The transition to 64-bit was the time to clean up the ISA (similar to what ARM did with AArch64).
If Intel had shown leadership, cleaned up the ISA instead of trying to stretch NetBurst to 10 Ghz and writing magic compilers that would make Itanium competitive, the world would be in a very different place right now. And so would Intel's shareholders.
name99 - Thursday, September 1, 2016 - link
Well. that's why my suggestion (and this is not backward looking, it was the suggestion I made when they first shipped Atom) is that Intel should have defined itself as "the kickass foundry company" not as "the x86 company".They pivoted once before, from memory to logic. But that was back in the days when they were run by engineers, not MBAs...
Meteor2 - Friday, September 2, 2016 - link
Intel makes a lot of money. They're not doing badly!TheinsanegamerN - Thursday, September 1, 2016 - link
So basically confirmed there will be no surface 4 then, unless MS brings windows RT back from the grave. thats sad. Not everybody can afford the surface pro lineJon Tseng - Thursday, September 1, 2016 - link
Yeah thats sad. So sequel to Cherry Trail.I have my Surface 3 running Dragon Age: Origins maxed out at 1080p yesterday. Ran just fine!!
ToTTenTranz - Thursday, September 1, 2016 - link
The Surface 4 could perfectly happen with a Pentium N4200.It's not the 1W difference from the Surface 3's X7 Z8700 that would prevent Microsoft from using a newer SoC.
TheinsanegamerN - Thursday, September 1, 2016 - link
firstly, yes it would. The pentium has a 6 watt TDP, with a SDP of4 watt. The surface 3's atom is a 2 watt chip. Doubling the TDP is not something many devices can handle without a massive overhaul on the design.The fact that the pentium also costs 5x as much is also a major issue. A jaguar is not a suitable replacement for a sentra, as it were.
TheinsanegamerN - Thursday, September 1, 2016 - link
*yes, it would be a problem. When is anandtech going to embrace the future that is 1999 with an edit button?Wolfpup - Thursday, September 1, 2016 - link
Which is why I wish a Tegra with Denver cores was allowed to run x86... :-/sonicmerlin - Monday, September 5, 2016 - link
I'm sure some Chinese oems will stuff Apollo trail into tablets.WorldWithoutMadness - Thursday, September 1, 2016 - link
so no more atom X[insert number]?What will the surface 4 be using then? core-m which is lotta expensive than atom?
Colin1497 - Thursday, September 1, 2016 - link
Aren't the NAS guys big customers on this end?Ryan Smith - Thursday, September 1, 2016 - link
With only 2 built-in SATA ports, it's not a great NAS SoC without additional SATA controllers.owan - Thursday, September 1, 2016 - link
Still, this is a chip class that we see in NAS units. I'm sure that adding an additional controller would be pretty trivial. May help explain why we see the 10W TDP... Intel might have looked at their volumes and found that the majority of this class of chip was shipping to devices with plugs (like NAS units) where the value of 2-3W power savings is low.ilt24 - Thursday, September 1, 2016 - link
The NAS manufactures use a wide range of processors, if you look at Synology for example, at the high-end they use Xeon E3's at the low-end they use ARM processors. I have an older version, but the Synology DS216+ uses a N3060 (Braswell).Colin1497 - Thursday, September 1, 2016 - link
Yeah, I know there's a range. I have a Rackstion 814+. The 814 was a Marvell SOC.fanofanand - Thursday, September 1, 2016 - link
Wow, so the tray price for an N4200 exceeds the retail price of an FX8320. Intel pricing is ludicrous.bill.rookard - Thursday, September 1, 2016 - link
That's because AMD doesn't have anything comparable. If they can scale their Zen architecture down into that space somehow, then you'd see the tray price drop.Just doing some math with some of Zens raw pre-production specs:
4c/8t = 66w, cut the cores down to 2c/4t = 33w, cut raw clockspeed down to half (remember, power usage isn't linear so you could wind up significantly lower) and you'd get down into the 16w range at least. Go with a low base clock of 1.5ghz and you could be in that 10w range or less... BUT we have to wait and see for a few more months
TheinsanegamerN - Thursday, September 1, 2016 - link
Binning may bring that down even more.ilt24 - Thursday, September 1, 2016 - link
OEMs don't pay anything near list price for Intel or AMDs processors.SpartyOn - Thursday, September 1, 2016 - link
As an owner of an embedded Braswell N3700 system, I think Intel should just give up in this space. Braswell was a massive failure; I swear my cheap-o Bay Trail Lenovo Yoga 2 w/ Windows tablet is just as fast and capable in terms of everyday use.On an AsRock N3700-ITX passively cooled board, the N3700 constantly throttles and can't maintain even passable GPU clocks - I think mine averages like 450 MHz when under stress. It's advertised as capable of 4K@30fps, which is laughable, when it struggles to even output 1080p streaming content without judder or frame skips every so often.
I should know - I tried hooking this up to my 4K LG OLED and the experience even navigating the mouse around the desktop was painful.
For a 6W chip in an actively cooled box, you would think that it would not need to throttle, but nay-nay. It's not like I'm asking it to do much, just stream some 1080p and maybe stream the occasional Steam In-home Network game.
This is the worst Intel chip I have ever bought, and I regret my purchase everyday.
patel21 - Thursday, September 1, 2016 - link
Hey you bought a 4K LG OLED TV, then you can upgrade your N3700 too,SpartyOn - Friday, September 2, 2016 - link
I sure can... the LG is actually hooked up to a GTX 1080 SLI system.This little box was just supposed to be for another TV.
Lolimaster - Thursday, September 1, 2016 - link
I don't think goldmont was a failure, the real thing is intel not evolving the performance of atom designs on purpose so they can maintain those ridiculous high prices on simple dual core core m/i5-i7u models.AMD Puma was a quad core with a $50 price tag, Cheetah was supposed to be closer to Phenom II/Athlon II in IPC.
Wolfpup - Thursday, September 1, 2016 - link
So Intel blocks Nvidia from releasing an x86 chip, and then doesn't even bother having their own tablet chip? Well, okay, they do with "Core-M", but obviously those cost as much as cheap tablets all on their own.I still think they made a mistake cutting out Nvidia, given more x86 is a good thing for Intel IMO, even if it's not an Intel chip.
And I'd much rather have Tegra Xx running real Windows than (eeeeew) Android.
name99 - Thursday, September 1, 2016 - link
"enabling Intel to reach cheaper markets and better compete with non-traditional devices based around ARM SoCs"Come on, Ryan, you don't have to engage in "Opinions differ on shape of the world" journalism. This thing is a PoS and you know it, just like your readers know it. The ONLY place it matters is in circumstances where x86 compatibility matters and performance does not --- maybe signage, industrial control, and scammy low-end Wintel PCs?
You really think this sad piece of junk, with performance at, what, half or less of an A72 and 3x the price, can compete against a decent Mediatek or Rockchip SoC, something like a Helio X20? Good luck with that.
aryonoco - Thursday, September 1, 2016 - link
I would love to see how these perform against Cortex A72 and A73. My guess, not very favourably.And I can guarantee that you can buy the fastest Cortex A72 from MediaTek and Rockchip for a lot less than these. As in, one sixth of the price of the cheapest Celeron here.
Sure, they are fabbed on 28nm, but they clock up to 2.4Ghz, so who cares. If a 28nm product can consistently beat a product on Intel's 14nm, what does that say about Atom's architecture?
x86 support you say? Well, how many of these units are going to be used for gaming? How many of them are going to be used for heavy applications like Photoshop? Not many. For everything else, general web browsing, office documents, media playback etc, Linux and ChromeOS will do just fine. And they work just as well on ARM (maybe even better seeing as ChromeOS on ARM will have better Android app compatibility).
Embedded systems pretty much all run Linux anyway. So do things like NASs. Routers started moving from MIPS to ARM a few years ago, a lot more of these devices will make the switch from x86 to ARM soon. And in fact looking at the Chromebooks that are going to launch in the second half of this year, that is exactly what is happening.
Not a pretty picture for Intel. Oh well, this will be Atom's last hurrah before being completely cancelled. At least Intel now has a license to fab ARM chips (hello XScale!)
yhselp - Thursday, September 1, 2016 - link
Does this mean it's now absolutely official that we won't see a proper successor to the Surface 3? I was really hoping for one. Seeing how the SoC in the Surface 3 costs $37 and draws 2 W (SDP), I can't see how it'd be possible to make a Surface 4. These new Atoms cost $107 and draw 4 W (SDP); the new Y-series cost $281 and draw about 2.33 W (SDP) at best if I'm not mistaken. The m3-7Y30 might do the job in TDP-down with a $250 price premium. That'd be tough to swallow even if they make a Surface 4. It'd be better than nothing, I guess...Arnulf - Friday, September 2, 2016 - link
" (ed: why couldn’t this be 3255?)"Because Down syndrome is a prerequisite for getting a job in IT marketing department.
Meteor2 - Friday, September 2, 2016 - link
That's a pretty crappy comment.fritsch - Monday, September 5, 2016 - link
It seems the HEVC-10 bit information for the BXT based chips (Apollo Lake) is wrong, the output of vainfo (linux hw acceleration lib) shows:vainfo: VA-API version: 0.39 (libva 1.7.1)
vainfo: Driver version: Intel i965 driver for Intel(R) Broxton - 1.7.1
vainfo: Supported profile and entrypoints
VAProfileMPEG2Simple : VAEntrypointVLD
VAProfileMPEG2Main : VAEntrypointVLD
VAProfileH264ConstrainedBaseline: VAEntrypointVLD
VAProfileH264ConstrainedBaseline: VAEntrypointEncSlice
VAProfileH264Main : VAEntrypointVLD
VAProfileH264Main : VAEntrypointEncSlice
VAProfileH264High : VAEntrypointVLD
VAProfileH264High : VAEntrypointEncSlice
VAProfileH264MultiviewHigh : VAEntrypointVLD
VAProfileH264MultiviewHigh : VAEntrypointEncSlice
VAProfileH264StereoHigh : VAEntrypointVLD
VAProfileH264StereoHigh : VAEntrypointEncSlice
VAProfileVC1Simple : VAEntrypointVLD
VAProfileVC1Main : VAEntrypointVLD
VAProfileVC1Advanced : VAEntrypointVLD
VAProfileNone : VAEntrypointVideoProc
VAProfileJPEGBaseline : VAEntrypointVLD
VAProfileJPEGBaseline : VAEntrypointEncPicture
VAProfileVP8Version0_3 : VAEntrypointVLD
VAProfileVP8Version0_3 : VAEntrypointEncSlice
VAProfileHEVCMain : VAEntrypointVLD
VAProfileHEVCMain : VAEntrypointEncSlice
VAProfileHEVCMain10 : VAEntrypointVLD
VAProfileVP9Profile0 : VAEntrypointVLD
Tralalak - Monday, March 13, 2017 - link
Compared to Atom (22nm Silvermont and die shrink 14nm Airmount), Goldmont is a bit of a beast. Both Atom and AMD’s Bobat/Jaguar/Puma core can fetch and issue up to two instructions. Goldmont can do three. Like AMD’s (or VIA’s) Goldmont has a full out of order execution engine. „Much of the changes with Goldmont were about improving the out-of-order execution compared to Silvermont, with a wider decoder, better branch prediction, and a larger out-of-order execution window. Goldmont can perform one load and one store per cycle, and it can execute up to three simple integer ALU operations per cycle. There’s new instruction support for hashing with SHA1 and SHA256, and there’s new support for the RDSEED instruction.“My Asus Aspire ES13 (ES1-332-P2VZ) with Intel Pentium Quad Core N4200 1.1GHz - 2.5GHz (Goldmont - Apollo Lake):
source: https://forums.anandtech.com/threads/intel-pentium...