I think most people truly don't understand what Atom is used for - in my opinion I believe one of purpose is for Intel to use as test bed for SOC in moving the whole system on single chip.
Technologies like Phi ( using many cores ) and Y and even U series come from this research.
I have one the newer Pentium based machines - it is significantly slower than other machines today - but similar speed to older machines like Core 2 and possibly my Surface Pro.
Speed is relevant to time it made - even the slowest atom will be faster that fastest machines 15 or 20 years ago.
It been a long time, like almost 25 years - but I used to do serious stuff with these Intel manuals - when they were actually physical pieces of paper. At first when I read this chart - I thought the Atom had AVX512 - I would like to see even the strongest ARM compete with that.
During my OS days, I remember going over to local Intel office in Atlanta with co-working and picking up Intel manuals for free because of my job - my favorite was the 386 Operating System writers guide - the last manual I got has paper was the sort of tan-gray Pentium manual - I had to purchase that one. I just pull out the original purple and black i486 manual next to Transputer manual from my book shelf in my bedroom. It is in pretty rough shape from my 1st job of seven years. We did get AMD manuals - not sure if they even have the specs on those now a days - back then the AMD chips were basically a clone of Intel chips - I not sure even CPUID could detect difference.
Phi has nothing to do with Atom. It was based on the GPU research that Intel did. Neither core design, nor manufacturing really relied on the research put into Atom.
And the fact that Atom is faster than a 20 years old CPU is pretty much irrelevant.
But what is relevant is the comparison to ARM. Since they both target the same low TDP space, the same usage scenarios, and more or less the same user base why wouldn't it be? There's a reason Atom based mobile devices never took off. They're either too slow in the low TDP, or have too high TDP in the higher performance models. And that's after Intel went through years of iterating and improving. If you look at what Apple managed to achieve with their A SoCs in just a few generations you start to see the limits and potential of each architecture or instruction set.
That first half is definitely not true. Phi is/was a combination of both. It was certainly originally based on the GPU research intel did, but the x86 core within the Knight's Landing line was unmistakably built off of Silvermont.
The original Larrabee had Pentium P54C-derived cores, which were carried through to the first Xeon Phi (Knights Corner). The second generation of Xeon Phi uses Silvermont Atom-derived cores. I don't know if they've said what Knights Mill uses, but it's sort of a half-generation, with some some hastily-added deep learning optimizations.
And before all of that (back in 2007) came Intel's 80-core TFLOPS research chip, which I this was reported to utilize some sort of RISC cores.
It started off from old Pentium cores and it started as GPU research. They gave up on the GPU dream because it was obviously not suited for that and decided to target the accelerator market.
Then they replaced the cores they were using with Atom cores but not because the Atom was a diamond in the mud. They did it because... "it was laying around" so to speak. :) Intel didn't have any other core that they could cram more than a handful of in one package and developing a whole new one made no sense. So they took the core they already had and that fit the bill: relatively low power, somewhat performant. But Phi itself wasn't designed with Atom in mind.
The Atom isn't a benchmark in any respect, unfortunately. It was the last pick on the dodge team when it came to Phi. And it's usually the last pick for anything since there are better options, ones with less compromises.
Atom with AVX512, aren't those all cancelled? I think the Phi's were the only CPUs where use of AVX2 or AVX512 doesn't cause a large slowdown of all cores. It's not a great thing if you have to actively avoid using AVX512 to get best performance...
Well at lot of Atoms have been cancelled which doesn't suggest at all it is on top.
The big issue with AVX512 is exactly that it is too inefficient, using way too much power, and that impacts performance quite badly. So a lot of software doesn't use AVX512 to avoid this slowdown.
I think a lot of software does not use AVX512 - because it so specialize and new - for what it does is significant - but of course if enabled it likely slow normal x86 operations down
I not sure how reliable your link is - others seem to discredit the information
It's a well-known fact that clock frequency goes down by a few hundred MHz whenever you use AVX2 or AVX512 even for a few instructions - and this slowdown lasts for long after the last use of such instructions. https://www.anandtech.com/show/11544/intel-skylake...
Where was it ever reported to appear outside of a Xeon Phi? I've never read such.
The cores in Xeon Phi are always customized to the application. The Silvermonts which had AVX512 also had 4-way hyperthreading (of which normal Silvermont has none).
People don't like Atom because it actually costs Intel almost nothing to manufacture a dual core i3 instead. Atom wouldn't exist anywhere, even in the low cost market, if Intel didn't want obscene margins. You don't need that kind of power efficiency in the laptop market for example.
One thing one forgets is that the atom is SOC and with i3 it takes addition chips to make the system, but as technology increases higher end models incorporate more of system - in fact the laptop I am typing this has 4.1Ghz quad core i7 with 20 Unit Vega GPU and 4G of HDM2 memory on a custom die. I believe EMIB will be the future.
To be honest, and just to clarify, I've never been prejudiced against Atom, even back in the netbook age and the single-core in-order models, the EeeBook I had worked okay on XP and was as expected, I took AnandTech's word on N3450 being okay and acceptable based on the Chuwi LapBook 14.1 review, and bought a Chuwi Hi13 with the same N3450 without doing additional research, and it turned out to be a massive disappointment in terms of performance. Perhaps it's because of the resolution of the screen, which is three times higher, but still - it's so cripplingly slow that's it's far, far from acceptable, and I barely use it because of that, and it wasn't that cheap either. So if Atom is to be a viable, albeit low-end, alternative, it definitely needs more, or my experience if because of the high-res screen.
On a more serious note, the cache control extensions are overdue.
I have been wanting CLWB (among other things) in x86, for more than a decade. It's useful when you modify some long-lived datastructure that you're not going to touch again (but might still read), for a long while. So you tell the CPU: "go ahead and write this out". There's nothing to be gained by deferring the write, and doing it now could reduce latency on the next cache miss in that set.
CLDEMOTE is a similar idea, but for shared data structures you know are going to be accessed by other threads. By proactively demoting it, you're reducing the time a thread running on another core will stall on it.
Another extension I wish they'd add is a pre-write instruction that allocates cache lines for a specified address range. That could let you avoid a fetch on write miss, when you're just going to overwrite the whole darn thing. For security reasons, the CPU would probably first have to zero it.
8 core A53s outperform most quad "big cores" from yesteryear. Snapdragon 625 for example, same single thread performance as the A15/Krait based SoCs, double the multithread performance, at a fraction of the cost.
Uh, no. Geekbench is lying to you on that one. When Geekbench starts treating x86 as a first class citizen and doesn't withhold AVX/2/512 optimisations when it has all of the ARM Neon ones, you'll see the world how it really is.
I wasn't referring to just Geekbench but a multitude of other benchmarks, emulators and real world experience. Also I never said anything about x86, I was merely defending the ARM 8 core .little philosophy which does in fact make sense
Nah. Bugger the BIG.little. Just have aggressive turbo states, so when cores are being underutilized... Funnel the TDP into boosting a few remaining cores.
Atom is dead and gone unless Intel can get Atom tablets selling again. I'm crazy enough to use a Cherry Trail Windows tablet as my main travel machine, it's fine for what it does and the battery life is excellent. It really is too bad that Intel dropped out of the 2W TDP segment.
The entire tablet market is in decline and has been for almost 2 years straight.
2 in 1's seem to have picked up some of the slack though, people are pretty content with the amount of capability on offer for something as simple as Facebook.
The entire tablet market isn't in decline. Per IDC, detachable tablets (which are mostly) windows based, started growing again in the second half of last year. I have started to notice a lot more people at work now moving from laptops to detachable tablets at work for their portables. My surface pro is a fantastic computer to have and I am unlikely to buy a laptop personally again.
Agreed with this. Detachables are a growth segment, but a lot of that growth is due to the fact that the segment enjoyed very few sales so even a comparably small numeric increase will seem like a substantial percentage growth rate. That said, I think the tablet market as a whole has reached a saturation point and further sales are generally going to customers purchasing replacements rather than customers entering the market for the first time. There's also a perception among consumers that there hardware, particularly ARM-based platforms like the iPad and various Android variants, aren't offering big enough improvements over previous generations that therefore justify a purchase. I'd say we're looking at stagnation and stability at the moment and suggest taking a wait-and-see approach before putting netbook-style nails into the tablet coffin.
There'll presumably be a bump this year from the new (very nicely priced) $330 iPad. There may also be bumps (or declines...) depending on whether Apple does or does not update the Pro line this year, and what they do with the mini (revamp it? cancel it?)
One thing that is interest is the MOVDIR* instructions, this deal with automously storing 64 bytes of memory fast - why would this be used - well 64 bytes is same thing as 512 bits - maybe used in AVX512 or possibly some kind of encryption algorithm. or a way to quickly past data between two cpu's in multi-cpu setup.
big.LITTLE becomes problematic when the two cores are not identical in various ways. One obvious way they need to be identical is the instruction set. Another less obvious way is in some of the cache details (eg protocol, cache line lengths).
But if there is one thing Intel loves, it is making incomprehensible minor differences to every damn CPU they ship. Which means it's not clear to me if they have ANY pair of cores that actually form a usefully matched big.LITTLE set.
Good question. Ars Technica asked about the security and related performance implications and received essentially no reply beyond "it's secure": "I pressed Intel Senior Principal Engineer Stephen Robinson pretty hard on the security aspects of this point—specifically, does Core-class branch prediction mean Core-class speculative execution vulnerabilities to exploits such as Spectre and Meltdown?
Robinson refused to speak with any detail on either Tremont's vulnerability to speculative execution attacks, or whether it would require software mitigations such as those necessary in Core class processors. He did not seem prepared for or comfortable with this line of questioning. Robinson also refused to comment on the performance implications of software mitigations if required..." https://arstechnica.com/gadgets/2019/10/intel-want...
One oddity in that document is that Cannon Lake is going to be picking up AVX-512. This is note worthy as the only Cannon Lake chips are going to be for mobile as Intel has gutted most of the Cannon Lake line up (desktop Cannon Lake has been dead for awhile). Putting a big fat power hungry SIMD unit into mobile parts seems weird when desktops don't yet have those ISA extensions.
... Atom is still a thing? Didnt they announce they were killing it a few years ago? Didn't the recent Cisco recall atom bug put a death nail in the platform? Don't get me wrong, I have always been a fan of Atom chips. Cheap small (all be it slow) x86 chips are a good thing in my book. But with the problems and lack of traction I thought Intel would have moved on to something else by now.
They also bought Altera (FPGAs) and several machine learning / computer vision chips that all feature hardwired (mostly ARMs and SPARC). Intel probably intends to replace some of those with Atom-family cores.
> They still sell Atom-based products into entry-level laptops,
"sell"? Now that's a bit of a stretch, as I'm pretty sure Intel were giving them away at $0 cost (or even less) in order to entice Shenzen manufacturers into using something other than ARM - it was always difficult to understand how the entire, finished tablet could retail at a price lower than the known Intel tray price of the CPU it contained. It's why the Intel mobile business lost $billions every year, as they actually "sold" squat, and just another example of Intel's extremely dodgy business practices.
You're conflating the entry-level laptop (AKA ChromeBook) market with their tablet SoCs.
Who can say why Apollo Lake and Gemini Lake are so prevalent in that market, today? I agree they're certainly not selling those chips at the listed tray price, but that doesn't necessarily mean they're taking a loss on them.
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.
56 Comments
Back to Article
mode_13h - Monday, April 23, 2018 - link
Cue the Atom haters and countless Atom vs. ARM specs wars and benchmark battles...HStewart - Monday, April 23, 2018 - link
I think most people truly don't understand what Atom is used for - in my opinion I believe one of purpose is for Intel to use as test bed for SOC in moving the whole system on single chip.Technologies like Phi ( using many cores ) and Y and even U series come from this research.
I have one the newer Pentium based machines - it is significantly slower than other machines today - but similar speed to older machines like Core 2 and possibly my Surface Pro.
Speed is relevant to time it made - even the slowest atom will be faster that fastest machines 15 or 20 years ago.
It been a long time, like almost 25 years - but I used to do serious stuff with these Intel manuals - when they were actually physical pieces of paper. At first when I read this chart - I thought the Atom had AVX512 - I would like to see even the strongest ARM compete with that.
mode_13h - Monday, April 23, 2018 - link
I miss the days when you could just call or email Intel and they'd send you a free set of bound, printed manuals. I still have my last set.HStewart - Monday, April 23, 2018 - link
During my OS days, I remember going over to local Intel office in Atlanta with co-working and picking up Intel manuals for free because of my job - my favorite was the 386 Operating System writers guide - the last manual I got has paper was the sort of tan-gray Pentium manual - I had to purchase that one. I just pull out the original purple and black i486 manual next to Transputer manual from my book shelf in my bedroom. It is in pretty rough shape from my 1st job of seven years. We did get AMD manuals - not sure if they even have the specs on those now a days - back then the AMD chips were basically a clone of Intel chips - I not sure even CPUID could detect difference.mode_13h - Monday, April 23, 2018 - link
I remember reading about Transputers. Fun to think about, back then. Years later, I had a professional encounter with (one of?) its descendent(s).mode_13h - Tuesday, April 24, 2018 - link
It's interesting to see that mesh architectures and many of the purported applications of Transputers (AI, etc.) have now become mainstream.close - Tuesday, April 24, 2018 - link
Phi has nothing to do with Atom. It was based on the GPU research that Intel did. Neither core design, nor manufacturing really relied on the research put into Atom.And the fact that Atom is faster than a 20 years old CPU is pretty much irrelevant.
But what is relevant is the comparison to ARM. Since they both target the same low TDP space, the same usage scenarios, and more or less the same user base why wouldn't it be? There's a reason Atom based mobile devices never took off. They're either too slow in the low TDP, or have too high TDP in the higher performance models. And that's after Intel went through years of iterating and improving. If you look at what Apple managed to achieve with their A SoCs in just a few generations you start to see the limits and potential of each architecture or instruction set.
Drumsticks - Tuesday, April 24, 2018 - link
That first half is definitely not true. Phi is/was a combination of both. It was certainly originally based on the GPU research intel did, but the x86 core within the Knight's Landing line was unmistakably built off of Silvermont.mode_13h - Tuesday, April 24, 2018 - link
You're both right.The original Larrabee had Pentium P54C-derived cores, which were carried through to the first Xeon Phi (Knights Corner). The second generation of Xeon Phi uses Silvermont Atom-derived cores. I don't know if they've said what Knights Mill uses, but it's sort of a half-generation, with some some hastily-added deep learning optimizations.
And before all of that (back in 2007) came Intel's 80-core TFLOPS research chip, which I this was reported to utilize some sort of RISC cores.
close - Wednesday, April 25, 2018 - link
It started off from old Pentium cores and it started as GPU research. They gave up on the GPU dream because it was obviously not suited for that and decided to target the accelerator market.Then they replaced the cores they were using with Atom cores but not because the Atom was a diamond in the mud. They did it because... "it was laying around" so to speak. :) Intel didn't have any other core that they could cram more than a handful of in one package and developing a whole new one made no sense. So they took the core they already had and that fit the bill: relatively low power, somewhat performant. But Phi itself wasn't designed with Atom in mind.
The Atom isn't a benchmark in any respect, unfortunately. It was the last pick on the dodge team when it came to Phi. And it's usually the last pick for anything since there are better options, ones with less compromises.
Wilco1 - Monday, April 23, 2018 - link
Atom with AVX512, aren't those all cancelled? I think the Phi's were the only CPUs where use of AVX2 or AVX512 doesn't cause a large slowdown of all cores. It's not a great thing if you have to actively avoid using AVX512 to get best performance...HStewart - Monday, April 23, 2018 - link
Well the stories of Atom's cancel must be internet lies - there is a lot of that going around especially when one is on top.On AVX512 it has a special need - AVX512 is new and it more likely that true performance comes when application uses AVX512 efficiently
Wilco1 - Tuesday, April 24, 2018 - link
Well at lot of Atoms have been cancelled which doesn't suggest at all it is on top.The big issue with AVX512 is exactly that it is too inefficient, using way too much power, and that impacts performance quite badly. So a lot of software doesn't use AVX512 to avoid this slowdown.
Note even minor use of AVX512 can reduce performance significantly: https://blog.cloudflare.com/on-the-dangers-of-inte...
HStewart - Tuesday, April 24, 2018 - link
I think a lot of software does not use AVX512 - because it so specialize and new - for what it does is significant - but of course if enabled it likely slow normal x86 operations downI not sure how reliable your link is - others seem to discredit the information
Wilco1 - Tuesday, April 24, 2018 - link
It's a well-known fact that clock frequency goes down by a few hundred MHz whenever you use AVX2 or AVX512 even for a few instructions - and this slowdown lasts for long after the last use of such instructions. https://www.anandtech.com/show/11544/intel-skylake...mode_13h - Monday, April 23, 2018 - link
Where was it ever reported to appear outside of a Xeon Phi? I've never read such.The cores in Xeon Phi are always customized to the application. The Silvermonts which had AVX512 also had 4-way hyperthreading (of which normal Silvermont has none).
Wilco1 - Tuesday, April 24, 2018 - link
Indeed, Xeon Phi uses an Atom derivative, and all new generations have been cancelled. See eg. https://www.theregister.co.uk/2017/11/16/intel_kil...Elstar - Tuesday, April 24, 2018 - link
Phi isn't dead. They just changed plans. "Knights Hill" is dead, long live "Knights Crest"Alistair - Monday, April 23, 2018 - link
People don't like Atom because it actually costs Intel almost nothing to manufacture a dual core i3 instead. Atom wouldn't exist anywhere, even in the low cost market, if Intel didn't want obscene margins. You don't need that kind of power efficiency in the laptop market for example.HStewart - Monday, April 23, 2018 - link
One thing one forgets is that the atom is SOC and with i3 it takes addition chips to make the system, but as technology increases higher end models incorporate more of system - in fact the laptop I am typing this has 4.1Ghz quad core i7 with 20 Unit Vega GPU and 4G of HDM2 memory on a custom die. I believe EMIB will be the future.rojer_31 - Tuesday, April 24, 2018 - link
Care to share the laptop model?Azurael - Tuesday, April 24, 2018 - link
Core-M (Y/U) parts are SoCs too, albeit MCM, so nothing gained there....Morawka - Monday, April 23, 2018 - link
Intel still unwilling to sell their Core architecture for under $100, thus they extend the life of Atom.MrSpadge - Tuesday, April 24, 2018 - link
Yeah, sure:https://geizhals.de/?cat=cpu1151&xf=820_1151
Celeron G3900 starts at 27€
Pentium Gold G4560 starts at 44€
yhselp - Wednesday, April 25, 2018 - link
To be honest, and just to clarify, I've never been prejudiced against Atom, even back in the netbook age and the single-core in-order models, the EeeBook I had worked okay on XP and was as expected, I took AnandTech's word on N3450 being okay and acceptable based on the Chuwi LapBook 14.1 review, and bought a Chuwi Hi13 with the same N3450 without doing additional research, and it turned out to be a massive disappointment in terms of performance. Perhaps it's because of the resolution of the screen, which is three times higher, but still - it's so cripplingly slow that's it's far, far from acceptable, and I barely use it because of that, and it wasn't that cheap either. So if Atom is to be a viable, albeit low-end, alternative, it definitely needs more, or my experience if because of the high-res screen.mode_13h - Thursday, April 26, 2018 - link
I'll bet it uses a single-channel DRAM configuration.mode_13h - Monday, April 23, 2018 - link
On a more serious note, the cache control extensions are overdue.I have been wanting CLWB (among other things) in x86, for more than a decade. It's useful when you modify some long-lived datastructure that you're not going to touch again (but might still read), for a long while. So you tell the CPU: "go ahead and write this out". There's nothing to be gained by deferring the write, and doing it now could reduce latency on the next cache miss in that set.
CLDEMOTE is a similar idea, but for shared data structures you know are going to be accessed by other threads. By proactively demoting it, you're reducing the time a thread running on another core will stall on it.
Another extension I wish they'd add is a pre-write instruction that allocates cache lines for a specified address range. That could let you avoid a fetch on write miss, when you're just going to overwrite the whole darn thing. For security reasons, the CPU would probably first have to zero it.
Elstar - Monday, April 23, 2018 - link
I wonder if any of the new Atoms will get a longer pipeline, or is that a premium feature reserved for the Core/Xeon line?StevoLincolnite - Monday, April 23, 2018 - link
I am hoping to see a 6-core Atom, would like to see Hex-core chips as the minimum going forward.Would be great in a low-powered 2-in-1 device for $300 or less.
DanNeely - Monday, April 23, 2018 - link
Unless they adopt something big.LITTLEesque I'd hope not. A 6 core atom makes as little sense as an 8 core A5x SoC ever has.Wardrive86 - Monday, April 23, 2018 - link
8 core A53s outperform most quad "big cores" from yesteryear. Snapdragon 625 for example, same single thread performance as the A15/Krait based SoCs, double the multithread performance, at a fraction of the cost.patrickjp93 - Monday, April 23, 2018 - link
Uh, no. Geekbench is lying to you on that one. When Geekbench starts treating x86 as a first class citizen and doesn't withhold AVX/2/512 optimisations when it has all of the ARM Neon ones, you'll see the world how it really is.Wilco1 - Tuesday, April 24, 2018 - link
Stop spreading lies. Geekbench supports x86 as a first class and optimizes for AVX2 and FMA. See https://www.geekbench.com/doc/geekbench4-cpu-workl... for proof. The truth is x86 is running out of steam.Wardrive86 - Tuesday, April 24, 2018 - link
I wasn't referring to just Geekbench but a multitude of other benchmarks, emulators and real world experience. Also I never said anything about x86, I was merely defending the ARM 8 core .little philosophy which does in fact make senseStevoLincolnite - Monday, April 23, 2018 - link
Nah. Bugger the BIG.little.Just have aggressive turbo states, so when cores are being underutilized... Funnel the TDP into boosting a few remaining cores.
HStewart - Monday, April 23, 2018 - link
There is actually a 16 core Atom server chip - it would be cool to see it 2in1 but they lack some Core chip functionalitymode_13h - Tuesday, April 24, 2018 - link
Not a minor point is that it burns too much power for such applications.rahvin - Tuesday, April 24, 2018 - link
There are already 16 core atoms, the C3000 chips that make up the Atom Server chips.serendip - Monday, April 23, 2018 - link
Atom is dead and gone unless Intel can get Atom tablets selling again. I'm crazy enough to use a Cherry Trail Windows tablet as my main travel machine, it's fine for what it does and the battery life is excellent. It really is too bad that Intel dropped out of the 2W TDP segment.patrickjp93 - Monday, April 23, 2018 - link
??? How about dumb terminals for Citrix in big corporations? Thousands of Atoms get sold for that purpose.StevoLincolnite - Monday, April 23, 2018 - link
The entire tablet market is in decline and has been for almost 2 years straight.2 in 1's seem to have picked up some of the slack though, people are pretty content with the amount of capability on offer for something as simple as Facebook.
Speedfriend - Tuesday, April 24, 2018 - link
The entire tablet market isn't in decline. Per IDC, detachable tablets (which are mostly) windows based, started growing again in the second half of last year. I have started to notice a lot more people at work now moving from laptops to detachable tablets at work for their portables. My surface pro is a fantastic computer to have and I am unlikely to buy a laptop personally again.PeachNCream - Tuesday, April 24, 2018 - link
Agreed with this. Detachables are a growth segment, but a lot of that growth is due to the fact that the segment enjoyed very few sales so even a comparably small numeric increase will seem like a substantial percentage growth rate. That said, I think the tablet market as a whole has reached a saturation point and further sales are generally going to customers purchasing replacements rather than customers entering the market for the first time. There's also a perception among consumers that there hardware, particularly ARM-based platforms like the iPad and various Android variants, aren't offering big enough improvements over previous generations that therefore justify a purchase. I'd say we're looking at stagnation and stability at the moment and suggest taking a wait-and-see approach before putting netbook-style nails into the tablet coffin.name99 - Tuesday, April 24, 2018 - link
iPad sales since 2015 are best characterized as flat, not declining.https://www.statista.com/statistics/269915/global-...
There'll presumably be a bump this year from the new (very nicely priced) $330 iPad.
There may also be bumps (or declines...) depending on whether Apple does or does not update the Pro line this year, and what they do with the mini (revamp it? cancel it?)
HStewart - Monday, April 23, 2018 - link
One thing that is interest is the MOVDIR* instructions, this deal with automously storing 64 bytes of memory fast - why would this be used - well 64 bytes is same thing as 512 bits - maybe used in AVX512 or possibly some kind of encryption algorithm. or a way to quickly past data between two cpu's in multi-cpu setup.mode_13h - Monday, April 23, 2018 - link
Or they just want to avoid a cache miss in the target by copying entire cachelines at a time.satai - Tuesday, April 24, 2018 - link
Can ve get big.LITTLE style design, where core architecture successor can trade some power efficency for some raw single-thread power?name99 - Tuesday, April 24, 2018 - link
big.LITTLE becomes problematic when the two cores are not identical in various ways.One obvious way they need to be identical is the instruction set. Another less obvious way is in some of the cache details (eg protocol, cache line lengths).
But if there is one thing Intel loves, it is making incomprehensible minor differences to every damn CPU they ship. Which means it's not clear to me if they have ANY pair of cores that actually form a usefully matched big.LITTLE set.
iwod - Tuesday, April 24, 2018 - link
The Goldmount + is getting very close to Sandy Bridge IPC, I am hoping this is another step closer.Does any one have figure for Goldmount + and Kaby Lake or Sandy Bridge transistor or die size?
dealcorn - Tuesday, April 24, 2018 - link
Is Tremont vulnerable to branch predictor issues?GreenReaper - Thursday, October 24, 2019 - link
Good question. Ars Technica asked about the security and related performance implications and received essentially no reply beyond "it's secure": "I pressed Intel Senior Principal Engineer Stephen Robinson pretty hard on the security aspects of this point—specifically, does Core-class branch prediction mean Core-class speculative execution vulnerabilities to exploits such as Spectre and Meltdown?Robinson refused to speak with any detail on either Tremont's vulnerability to speculative execution attacks, or whether it would require software mitigations such as those necessary in Core class processors. He did not seem prepared for or comfortable with this line of questioning. Robinson also refused to comment on the performance implications of software mitigations if required..."
https://arstechnica.com/gadgets/2019/10/intel-want...
Kevin G - Tuesday, April 24, 2018 - link
One oddity in that document is that Cannon Lake is going to be picking up AVX-512. This is note worthy as the only Cannon Lake chips are going to be for mobile as Intel has gutted most of the Cannon Lake line up (desktop Cannon Lake has been dead for awhile). Putting a big fat power hungry SIMD unit into mobile parts seems weird when desktops don't yet have those ISA extensions.CaedenV - Tuesday, April 24, 2018 - link
... Atom is still a thing? Didnt they announce they were killing it a few years ago? Didn't the recent Cisco recall atom bug put a death nail in the platform?Don't get me wrong, I have always been a fan of Atom chips. Cheap small (all be it slow) x86 chips are a good thing in my book. But with the problems and lack of traction I thought Intel would have moved on to something else by now.
mode_13h - Tuesday, April 24, 2018 - link
No, they killed off their cell phone SoC products, which were built around the Atom-series cores.They still sell Atom-based products into entry-level laptops, embedded applications, low-power servers, and HPC (Xeon Phi).
https://ark.intel.com/products/codename/80644/Apol...
https://ark.intel.com/products/codename/83915/Gemi...
https://ark.intel.com/products/series/97941/Intel-...
https://ark.intel.com/products/series/87465/Intel-...
https://ark.intel.com/products/series/132784/Intel...
They also bought Altera (FPGAs) and several machine learning / computer vision chips that all feature hardwired (mostly ARMs and SPARC). Intel probably intends to replace some of those with Atom-family cores.
CityBlue - Wednesday, April 25, 2018 - link
> They still sell Atom-based products into entry-level laptops,"sell"? Now that's a bit of a stretch, as I'm pretty sure Intel were giving them away at $0 cost (or even less) in order to entice Shenzen manufacturers into using something other than ARM - it was always difficult to understand how the entire, finished tablet could retail at a price lower than the known Intel tray price of the CPU it contained. It's why the Intel mobile business lost $billions every year, as they actually "sold" squat, and just another example of Intel's extremely dodgy business practices.
mode_13h - Thursday, April 26, 2018 - link
You're conflating the entry-level laptop (AKA ChromeBook) market with their tablet SoCs.Who can say why Apollo Lake and Gemini Lake are so prevalent in that market, today? I agree they're certainly not selling those chips at the listed tray price, but that doesn't necessarily mean they're taking a loss on them.