The Intel Atom Thread

[Arachnotronic]

Intel Atom 'Broxton' - Geekbench & GFXBench: First Results

Some early Broxton results, after such a long wait. The chip is being recognized as a 2.3GHz Pentium II/III by Geekbench. Assuming 2.3 GHz clockspeed I find the results rather promissing. Not on the same level as high-end ARM SoCs, but already >30% faster than an average Atom x7-Z8700 (not bad for an ES):

- Geekbench 3
Single-core: 1301
Multi-core: 4385

https://browser.primatelabs.com/geekbench3/6034406

*Faster than most AMD A8-7410 (fastest Carrizo-L, 2.2-2.5 GHz, 15W TDP) devices out there.
*Highest score for Stoney Ridge ES (6W TDP?) = 1480 ST / 2417 MT.


- GFXBench
Car Chase Onscreen: 35.7 FPS

https://gfxbench.com/device.jsp?D=I...&testgroup=graphics&benchmark=gfx40&var=score

Pathetic showing. Intel has barely caught up with the A57 inside of the S6.

https://browser.primatelabs.com/geekbench3/compare/2683519?baseline=6034406

 

[Nothingness]

Well it's not bad for sure, against one of the best x7-Z8700 results:
http://browser.primatelabs.com/geekbench3/compare/2517688?baseline=6034406

I wonder if Intel fans will still complain about GB cryptographic bias

The comparison looks bad: remove AES and SHA1 from the integer part of the test and see what you get... FP results are odd, some are very good, the others no. The issue seems to be memory. Let's wait for a non-ES chip.

Pathetic showing. Intel has barely caught up with the A57 inside of the S6.

https://browser.primatelabs.com/geekbench3/compare/2683519?baseline=6034406

Broxton crypto looks good, hmm

 

[Sweepr]

Pathetic showing. Intel has barely caught up with the A57 inside of the S6.

https://browser.primatelabs.com/geekbench3/compare/2683519?baseline=6034406

They've long ago given up on high-performance phones / tablets with Atom. >30% better CPU performance and a beefier Gen 9 iGPU would do wonders for entry-level $100-300 x86 devices. Let's see where final chips will land.

The announcement of Apollo Lake was made by Navin Shenoy, corporate vice president and general manager of Intel’s Client Computing Group, during his keynote at the Intel Developer Forum. He said that the company has already over 100 design wins based on the Apollo Lake processors, but didn’t reveal any technical specifics, only that Apollo Lake follows Cherry Trail (Atom X5 series) and Braswell (Celeron and Pentium) products.

BTW +100 designs wins for Apollo Lake already. Early Q3 launch?

www.digitaltrends.com/computing/apollo-lake-intel/#ixzz45pRojxZ0

 

[ehume]

Lovely! HP will put this in their cheapest laptops, and I can replace my old HP Stream 11 with a new HP Stream 13! (You can tell I'm bottom-feeding here.)

Looking at those links. Does Lenovo really have a model named the Invalid? What next? The Lenovo Wheelchair?

 

[IntelUser2000]

- Geekbench 3
Single-core: 1301
Multi-core: 4385

https://browser.primatelabs.com/geekbench3/6034406

No wonder they reduced their mobile expenditures. It looks like looking at individual scores, its about 20% faster. It should be 25-30% with dual channel memory.

Even then, that's quite horrible. They may be artificially crippling Atom to keep Core M look better than it is.

Few years ago Mr. Skaugen(when he was in the server division) said that they wouldn't cripple a line to push another one. He was obviously talking about Itanium, but said if Atom was the architecture to go, he'd push that, and not cripple it.

They should have done whatever it is required to DOUBLE CPU performance and increase GPU performance by 6x. Then the CPU would be barely competitive, but GPU would be on par with next gen ARM parts.

Pathetic showing. Intel has barely caught up with the A57 inside of the S6.

I had the impression for quite a while that Intel is forever behind in architecture, which is why they needed +1 process advantage, cancelling them out. Silverthorne = A9, Silvermont = A15, Goldmont = A57? It looks like Goldmont isn't even 14nm A57, but a regular A57.

 

[Shivansps]

Ok. That is what I waited a lot.

Finally seems that Apollo Lake is getting the performance of Core 2 Conroe. If that is true, seems that ARM will make their massive improvements because Intel is catching up.

BTW.. Stoney Ridge is a Dual Core. And ST wise defeats the Apollo Lake.

It looks like Stoney Ridge beat it in FP perf.

 

[Shivansps]

From it what looks like, it seems to be on pair with Pentium E2140 ST perf, what is not that impresive considering E2140 is a 1.6Ghz part, a E5200 will just smash it.

 

[jpiniero]

They should have done whatever it is required to DOUBLE CPU performance and increase GPU performance by 6x. Then the CPU would be barely competitive, but GPU would be on par with next gen ARM parts.

So, basically Core M... except with half the power draw.

 

[Sweepr]

If they can hit 1500 ST/4500-5000 MT it will be a meaningful upgrade while not closing the gap with Kabylake-Y too much. Personally I still care about small x86 cores and I'm looking forward to Apollo Lake & Broxton.

Like it or not, Core-based Celeron/Pentium notebooks are becoming rare these days, so this is what you're going to find inside Chromebooks & Windows-based Cloudbooks in the next few years.

 

[dark zero]

Lovely! HP will put this in their cheapest laptops, and I can replace my old HP Stream 11 with a new HP Stream 13! (You can tell I'm bottom-feeding here.)

Looking at those links. Does Lenovo really have a model named the Invalid? What next? The Lenovo Wheelchair?

INBF HP releases the HP Stream 11, 13 and 14. The latter one using a E2 Stoney Ridge with better ST performance compared to Apollo :V

 

[VengenceIsMine]

Anyone have any sense of how much the awful memory scores for the Broxton is handicapping other results? The difference in memory perf is pretty huge given that Surface 3/CT is a close to year old design vs something that won't even ship in quantity for another 3-4 months at the earliest.

 

[Sweepr]

AnandTech's article about Apollo Lake:

Intel Unveils New Low-Cost PC Platform: Apollo Lake with 14nm Goldmont Cores

This week, at IDF Shenzhen, Intel has formally introduced its Apollo Lake platform for the next generation of Atom-based notebook SoCs. The platform will feature a new x86 microarchitecture as well as a new-generation graphics core for increased performance. Intel’s Apollo Lake is aimed at affordable all-in-ones, miniature PCs, hybrid devices, notebooks and tablet PCs in the second half of this year.

...Traditionally, Intel discusses options and its vision, but not actual PCs, at its IDF trade-shows. It remains to be seen whether PC makers decide to build low-cost convertibles or ultra-thin notebook designs, or will stick to more traditional clamshell notebooks. In fact, we will learn more about BTS (back-to-school) plans of major PC OEMs regarding Apollo Lake at the upcoming Computex trade-show in early June. There's also IDF San Francisco in August where Intel may open some lids on how the new Goldmont core differs from Airmont.

www.anandtech.com/show/10256/intel-unveils-apollo-lake-14nm-goldmont

Apollo Lake design wins at Computex 2016.

 

[ninaholic37]

Why Your Next Laptop Should Be a 2 in 1

"When compared to traditional laptops, a detachable keyboard or 360-degree hinge makes everything better. There's nothing a traditional laptop can do that a similarly priced 2 in 1 can't do."

What a crock. There's no such thing as a 2-in-1 $199 HP Stream 11, and if there was it would probably be trash compared to the similarly priced laptop version. Take your Frankenstein 2-in-1s and shove them up your you know where Intel!

 

[nvgpu]

http://asrock.com/mb/Intel/J3710M/index.us.asp

http://asrock.com/mb/Intel/J3710-ITX/index.us.asp

http://ark.intel.com/products/codename/66094/Braswell#@Desktop

 

[IntelUser2000]

If they can hit 1500 ST/4500-5000 MT it will be a meaningful upgrade while not closing the gap with Kabylake-Y too much. Personally I still care about small x86 cores and I'm looking forward to Apollo Lake & Broxton.

They need to either abandon that mentality or create a large core. Current Core chips are medium-sized cores. Get out a new core with 1024 ROBs and 8-10 wide execution. That's how they can have 28 of them in Skylake EP and get it under 4.5W with Core M. If they are at a point where they really can't get those chips out because we've completely exhausted ILP extraction...

well, they'll do business for 10-15 more years and close shop. I think at certain point they'll have to face a time when the PC sales are so small that they CAN sacrifice everything for mobile.

So, basically Core M... except with half the power draw.

Well, similar CPU, but 50% faster GPU. Core M GPU isn't enough. And that should have been last year.

 

[Nothingness]

Choosing 8 to 10 instructions in an instruction window of 1024 is impossible while maintaining frequency (and without killing power due to wrong speculation that will explode).

 

[Arachnotronic]

They need to either abandon that mentality or create a large core. Current Core chips are medium-sized cores. Get out a new core with 1024 ROBs and 8-10 wide execution. That's how they can have 28 of them in Skylake EP and get it under 4.5W with Core M. If they are at a point where they really can't get those chips out because we've completely exhausted ILP extraction...

well, they'll do business for 10-15 more years and close shop. I think at certain point they'll have to face a time when the PC sales are so small that they CAN sacrifice everything for mobile.

Well, similar CPU, but 50% faster GPU. Core M GPU isn't enough. And that should have been last year.

The cores inside of SKL-EP aren't even the same as the ones inside of laptop/mainstream desktop Skylake.

 

[IntelUser2000]

Choosing 8 to 10 instructions in an instruction window of 1024 is impossible while maintaining frequency (and without killing power due to wrong speculation that will explode).

That's the issue they have. They want to keep those margins, and before the mobile explosion, the pricing was justified(at least somewhat) because they had that performance difference. Bottom end Intel chips were probably 10x faster than the fastest ARM based ones.

Now the justification for their pricing is because there's no alternative in the Windows space and the only higher performing chips are pricey Core ones. People didn't have that thinking when Medfield was out. They were expecting that they'd be #1 in mobile space or at least #2 and performance improvement of ARM competition would level out short time after. Then Apple came with A7, forcing everyone to admit that Intel would have zero chance going into Apple devices. They'd reason "oh they'll be the next Qualcomm". And they'd say it doesn't matter Apple chips are better since "its not direct competition".

Now, non-Apple top end ARM chip performs almost as well as the Core M chip(if not there already depending on how do you view Intel C compilers), with cost, battery life, and form factor of an Atom chip/device

I can't help but admit that Intel is in a serious quandry here. They want to satisfy shareholders and keep those massive revenues/margins, they want to be dominant in mobile, yet they can't do much about it. Ideally Atom would perform as well as Apple A9 chips(or current Core M), and regular Core would be still 2x above that. But you run into "real world" issues like you are pointing out how unrealistic it is expecting such big cores to happen. They need a "big core" chip not just because they want to improve PC sales, but to keep the pricing/margins and allow lower cores like Atom to grow so they can be competitive in mobile(I know its not just CPU, but the entire SoC is behind).

Whether they aren't capable of doing it due to valid reasons like not being able to make cores that perform another level above Core, or they don't want to do it due to financial reasons, they are in big trouble. Technical troubles will eventually manifest themselves financially. Perhaps their time is past though, just like disk/cassette rental places.

 

[VirtualLarry]

That's the issue they have. They want to keep those margins, and before the mobile explosion, the pricing was justified(at least somewhat) because they had that performance difference. Bottom end Intel chips were probably 10x faster than the fastest ARM based ones.

Now the justification for their pricing is because there's no alternative in the Windows space and the only higher performing chips are pricey Core ones. People didn't have that thinking when Medfield was out. They were expecting that they'd be #1 in mobile space or at least #2 and performance improvement of ARM competition would level out short time after. Then Apple came with A7, forcing everyone to admit that Intel would have zero chance going into Apple devices. They'd reason "oh they'll be the next Qualcomm". And they'd say it doesn't matter Apple chips are better since "its not direct competition".

Now, non-Apple top end ARM chip performs almost as well as the Core M chip(if not there already depending on how do you view Intel C compilers), with cost, battery life, and form factor of an Atom chip/device

I can't help but admit that Intel is in a serious quandry here. They want to satisfy shareholders and keep those massive revenues/margins, they want to be dominant in mobile, yet they can't do much about it. Ideally Atom would perform as well as Apple A9 chips(or current Core M), and regular Core would be still 2x above that. But you run into "real world" issues like you are pointing out how unrealistic it is expecting such big cores to happen. They need a "big core" chip not just because they want to improve PC sales, but to keep the pricing/margins and allow lower cores like Atom to grow so they can be competitive in mobile(I know its not just CPU, but the entire SoC is behind).

Whether they aren't capable of doing it due to valid reasons like not being able to make cores that perform another level above Core, or they don't want to do it due to financial reasons, they are in big trouble. Technical troubles will eventually manifest themselves financially. Perhaps their time is past though, just like disk/cassette rental places.

Maybe Intel's entire marketshare, has always been somewhat of a "house of cards", relative to their competition, propped up by biased / influenced / sponsored benchmarks?

They have a long and sordid history in that dept.

Maybe now, the true performance of Intel is just being known?

 

[coercitiv]

Like it or not, Core-based Celeron/Pentium notebooks are becoming rare these days, so this is what you're going to find inside Chromebooks & Windows-based Cloudbooks in the next few years.

They get maybe one or two more years before decent (compelling) ARM based hybrids hit the shelves, from both Apple and Google, with both iOS and Android teams laser focused on improving multitasking user experience. The only way they can still grow their business is to carve deeper into PC territory, and that big PC monster is so slow that it barely registered the mobile piranhas tasting it's tail.

This will be an epic battle, and I'm pretty sure we're going to see a far more agile Intel in the years to come.

Maybe now, the true performance of Intel is just being known?

First of all, be careful on how you interpret performance benchmarks for mobile chips. Just like in the past people fell for Intel optimized benchmarks, now they may just as well fall for benchmarks that are not truly representative of SoC performance. That's your house of cards right there, or at least part of it.

Second, Apple is selling you similarly priced ARM based products with the best and most expensive PC equivalents. The iPad Pro is asking for the same $$$ as the Surface Pro 4. If Intel's perf/dollar ratio is bad, who has the better one?

Third, watch this video from 11:30 to 14:00. Intel made the wrong bet on tech ever since the beginning of this century, and it's still not fully clear to me whether they are fully committed to a mobile strategy. On the surface all they do & talk is mobile, bringing down to passive cooling their high performance cores, optimizing future processes for even more perf/w, investing in the future IoT etc, but when I look at the way they handled their mobile strategy I see leadership who still thinks they can use old tactics, pull an ace from their sleeve and steam through the competition in order to get back asap to the old status quo (the peace and quiet of high margins).

They lost smartphones, they lost cheap media consumption tablets, now it's gonna be about high end tablets/hybrids with productivity in mind. Until now we saw Intel losing fights in enemy lands, now we're seeing skirmishes at the border and soon enough, as I wrote at the beginning of this post, we'll see the battle move in Intel's territory.

Some think Intel is losing the performance crown, but I beg to differ. I think Intel currently has the better tech (performance wise), however I also fear this is exactly the reason they're still too confident and not taking the bull by the horns. Being comfortable wearing the performance crown is surprisingly the reason they allow this siege to take place, and I'm not so sure their walls are that high and that thick. And since I made so many analogies so far, here's a song from a PC game that perfectly describes this scenario.

On the other hand, this is the nature of the beast: it's easy to watch mistakes being made from afar, another thing entirely to explain investors why you sacrifice today's profits to build strength for a storm that may or may not come several years later.

 

[ksec]

https://browser.primatelabs.com/geekbench3/compare/6110477?baseline=6034406

Comparing to a 5 1/2 years old SandyBridge CPU, it is about the same in Interger, awful in FP and Memory. I am thinking it may not be using all the SSE instruction it can, Non ES Sample should improve. But memory is likely due to 2 vs 1 Channel.

I am waiting to see how it would work out, my SandyBirdge CPU + 8GB Ram + old SSD works wonderfully well most of the time for office and Internet usage. I get a few crashes from GPU drivers, which should be hugely improve with Intel Gen 9 graphics + much more mature drivers.

I a mostly missing faster network such 802.11ac, faster USB.

Other then that I have yet to see a major reason to upgrade the office computer. It would be nice if Intel offer all these in a smaller package and Windows 10. That would definitely make upgrading much more attractive.

The problem is again, Lenovo, HP, Dell the three biggest Business vendor sill do not include SSD as standard.

 

[CHADBOGA]

Maybe Intel's entire marketshare, has always been somewhat of a "house of cards", relative to their competition, propped up by biased / influenced / sponsored benchmarks?

They have a long and sordid history in that dept.

Maybe now, the true performance of Intel is just being known?

Yeah if you ignore the experience of everyone who has ever bought their CPU's. D:

 

[VirtualLarry]

Yeah if you ignore the experience of everyone who has ever bought their CPU's. D:

Like a Celeron D, you mean?
Edit: Nevermind, this is the Atom thread, those processors speak for themselves, no need to bring up Intel's other poor performers.

 

[dark zero]

Like a Celeron D, you mean?
Edit: Nevermind, this is the Atom thread, those processors speak for themselves, no need to bring up Intel's other poor performers.

There is a need. Edinson or Quark... ARM processors gets better performance than them.

 

[mikk]

Well it's not bad for sure, against one of the best x7-Z8700 results:
http://browser.primatelabs.com/geekbench3/compare/2517688?baseline=6034406

That's unfair because you are comparing a x64 Geekbench result from Z8700 against a x32 result from Broxton. You should compare 32 bit results from Z8700 and to make it more comparable looking for results with similar memory scores because this Broxton is obviously running in singlechannel mode.