You all almost sound like this is Qualcomm's first go at making a mobile processor. I'd find it very hard to believe that they'd not take into consideration power consumption and heat or any other negative factors they considered when they made their other chips. Having said that, all that could be out the door in the race bigger and better, who knows. Guess we all will have to wait and see how this things performs in an actual smartphone.
How does this compare to the Ivybridge or Haswell Y series processors in terms of CPU/GPU performance? Would like to see those comparisons. So we can extrapolate how this will compare to a fanless Core product from Intel. The gap is there in terms of connectivity but the perf gaps on CPU/GPU would be interesting to note.
Ivy Y no longer seems interesting to me but for the moment, Haswell 15W (both HD4600/5000) would be super interesting to add to the "great equalizer" page of charts. Doesn't Anand have an Acer with i7-4500U as well as the MBA with i5-4250U? Demand he run some benchmarks for you so you can update those charts. The comparison to i5-3317U made sense against APQ8064T, but 8974's contemporary is Haswell, not Ivy.
And when you have Haswell Y samples, those would be neat, too. :)
I wonder how Rouge is going to compare to this. My guess is Apple will ship a rouge powered iPhone and iPad before devices start shipping with Qualcomm.
From what I'm seeing here this an amazing chip overall. That GPU really surprised me but I'm gonna wait till the battery life/power draw can be tested.
I'm not sure you could call them the Intel of Mobile. There's a couple of key differences - the first and most obvious being that Intel is leading in x86 IPC, power management and, this is the big difference, manufacturing. Intel won't and can't go away because they have a tangible advantage over their competitors that will require incredible investment to overcome.
The other is that Qualcomm have a large slice of an ever-increasing pie, but Intel have basically the whole x86 cake to themselves, with the crumbs falling to AMD.
It's probably also worth mentioning that Qualcomm seem better at producing the right improvements for consumers, wheras Intel were late to the low-power party and slow to get going on graphics to (although both are clear foci of the company now).
Interestingly, I believe Qualcomm's mobile division is basically AMD/ATI's old mobile division, which they didn't think was worth it to keep. That's looking a like a pretty huge mistake right now.
Thanks for the grammar correction, teach. Now that I know it any way you can teach it to my phones keyboard? Grammar is useful for helping to avoid ambiguity but the previous could only be understood one way. IOW, your correction wasn't needed unless you were just being anal.
You make some good points and it's clear that not clear what it means when someone says "X is the Intel of Y". Intel using their fab is certainly a part of their success. That they are THE key x86 player is, I think, the more important touchstone. It was that which I was speaking of when I spoke of them. I THINK they are, by far, the biggest, most profitable supplier of ARM based SoC in the world. Samsung is huge but they consume what they make. TI is leaving the mobile field and haven't had a major design win for awhile. BTW you're right about ATI and adreno but the 300+ series is vastly different from the architectures Qualcomm inherited from ATI.
Very inconclusive without any power measurements. In its tablet reference form factor certainly beats modern SoCs in phone form factors. I'd like to see it against Tegra 4, Bay Trail, and Haswell (4.5W SDP) in a tablet form factor.
I agree, I want to see it next to Tegra 4, Bay Trail, and also Kabini and Temash.
But I have to say, it looks good, really good. Especially that GPU.
To think that Adreno was AMD/ATI's embedded graphics division that AMD didn't think was worth anything and sold for peanuts (something like $30M IIRC) to Qualcomm. Dirk Meyer really ruined that company.
The engadget has the benchmark for snapdragon on phone and tablet. It can be seen that the snapdragon 800 is very fast in tablet setup, but only little bit faster than snapdragon 600 in phone setup. I guess this partly answered your TDP concern.
I Love the Equalizer Part. Finally we have some sense of how far we are from Mainstream Desktop Performance. Would be great if Haswell were added to the Chart, and a UL Haswell running at similar Power usage in there as well. ( Although i believe Intel would properly pressure Anand not to have done so )
Can't believe how far behind Atom's GPU is at this point and how long Intel sandbagged the market with it...
I've been rather disconnected from the Android news/rumor mill, any chance Snapdragon makes it into the next small Nexus tablet design? I really wanna downsize/upgrade from my OG Transformer...
That would be awesome, but I highly doubt it. Unless they radically change their target market for that device, I don't think you'll see a bleeding edge SOC or super high resolution display. It was meant to win back the market share lost to the likes of the budget friendly Kindle/Nook tablets who were eating Google's lunch in the tablet market while using their OS at the core. I just don't see a Snapdragon 800 in a ~$200 device, for the same reasons you don't see Swift or Retina displays in the iPad Mini. I think we're much more likely to see an S4 Pro quad/Adreno 320 (possibly Snapdragon 600?). Even that would be a hell of an upgrade to the existing Tegra 3, and I'd happily throw my Nexus 7 up on eBay for that. Admittedly, I may end up with an x86 Windows 8.1 tablet with Silvermont or Temesh instead of the next Nexus 7 anyway.
He's comparing what he has (Snapdragon development tablet) with other things that he has.
Clock frequency is pointless to compare by itself. Sure A15 might be faster clock-for-clock, but it's also more power hungry so you won't be seeing it hit 2.0 Ghz anytime soon. Looks like a Snapdragon 800 can easily clock higher which more than compensates its IPC disadvantage compared to A15 while keeping power usage in check.
Qualcomm is sitting very comfortable right now. No wonder their market cap is now bigger than Intel, while they have a fraction of Intel's revenue/profits. The market clearly likes what they see from them.
A15 is already at 1.9GHz with Tegra 4. If A15 was built on 28nm HPM then it would also reach 2.3-2.5GHz. I bet we'll get something like that this year.
Note there is no evidence A15 uses more power. In S4 tests it seems the Exynos Octa has lower power consumption than Krait.
Tegra 4 finally has a fast GPU - nobody has benchmarked it yet but the one available score (Egypt HD at 57fps) suggests it will be a close match with Adreno.
Tegra 4 is already A15 on HPM and it is capped at 1.9Ghz, no evidence on how much further it can be pushed. Also Exynos Octa have 4 A7s to reduce power consumption, in GSMArena's benches it still fall behind S600 in battery life.
The Exynos variant of S4 does indeed fall behind in battery life in the GSMArena test, but that was on talk time and video playback, ie. not related at all to the CPU. QC has a big lead in low-power modems and it shows. Browsing was within 6%, so that confirms that the A15 uses similar power as Krait to get slightly better performance. I'm still hoping Anand will do a full review...
Yes, the architecture comparison between A15 and Krait 400 and next-gen Swift will be very interesting.
Add in Bay Trail and Temash and we CPU enthusiasts are really experiencing a golden age. When was the last time we had so many different competing designs and architectures? When you compare the progress from Scorpion to Krait 200 to Krait 400 (or A9 to A15) in a matter of 2 years... the pace of development in the mobile world is astonishing.
I'm also relying on Anand to do some full analysis and review of all these architectures once they become available. It's pretty clear that Tegra 4 is going to be cheaper than Snapdragon 800, it would be great to see them side by side, with power analysis and all.
"The overall graphics score from Adreno is amazing. We're looking at almost 2x the next fastest contender here, the Adreno 320 based Snapdragon 600."
Why is that so amazing? Adreno 320 came out first last fall, which is about a year before S800 with Adreno 330 comes out. Are you saying you're surprised mobile GPU's can double in performance every 12 months?
In the great equalizer charts, can someone please explain how GeForce 8500 manages to outperform the 7900 GTX in DXBenchmark 2.7? The 8500 had similar performance to a 7600GS from what i remember.
Where will the new generation Atom show up in the list? As I understand it, it will have graphics comparable to the HD4000 in the Ivy Bridge, so in theory it should outperform this Qualcomm chip?
No, it will have graphics based upon the HD4000 architecture - but it will have fewer units and run at a lower speed, in order to attain TDP requirements.
It seems to me that everything is depending on the power budget, as the dual core A15 inside Nexus 10 is faster in some tests then the quad A15 in S4. Or I am wrong ?
Yes in some tests (like 3dmark physics) it is very obvious the quad A15 aren't running anywhere close to their max frequency. This is not surprising last time I looked at some power draw figures just two A15 cores at max frequency alone (so without graphics) could already exceed the power budget of the whole chip. So we'll see how that new Snapdragon chip compares once it's inside a smartphone. Should still be plenty fast but some scores might suffer (maybe we're going again to see freezer vs. no freezer results :-)).
I just wanted to make people realize something:Assassin's creed 1 & 2 run very well on an old x1950pro, and this chip seems to me that it's faster. When they said that soon mobile platform would exceed the(actual but) old consolle generation they weren't kidding.
This looks great - but frankly I need to see this vs Tegra 4 and A7. Also - I agree we need to see vs HD5000/HD5200 Haswell.
Tegra 4 is probably a toss up performance wise - and will come down to power (which Qualcom normally wins). Kinda makes me sorry I pre-ordered a Shield when this will be in the Note 3 and I could just attach a controller. Hopefully Nvidia can do a better job of getting AAA games to actually support Shield. I'm sick of all the blockbusters being iOS only (or 6 months early).
Apple seems to double performance every year, and with PowerVR 6 - you could get a bigger bump. That would easily surpass this. If AAPL would just come out with a larger screen, I'd go back to iOS this time around.
And IF we see Haswell tablets with 7-9 hours of battery life - I'm not sure all of this stuff doesn't get relegated to the dust bin of history with Intel taking 90%+ of the tablet/hybrid market.
Anyone have any info on Saltwell's GPU? I love my Latitude 10's ability to run full Windows programs - particularly Office. The CPU beats anything in this article and the power use is low enough I get 16 hours of battery life. But the GPU is so far behind all of these other chips even a 2x-3x improvement isn't going to make it competitive. Why can't they throw in whatever Apple is going to use in the A7. They own part of PowerVR too!
I'll be interested to see how the actively-cooled Tegra 4 in Shield performs compared to the standard passively-cooled stuff. I'd suspect that if you're going to be gaming for a while, the Shield may pull away. Tablets and smartphones these days aren't designed to run at full tilt all the time.
The most interesting thing about this generation of SoCs to me is the CPU performance. I don't game all that often, so I'd much prefer a device that has CPU performance to spare. It's the opposite of the problem that we had on the desktop years ago. Now we have GPU performance sufficient for UI acceleration, media decoding, and mainstream gaming, but we still have CPU performance that holds the platform back.
No with a processor offering 4x the performance of anything that ARM offers. Believe it or not, most corporate users want something that can run enterprise level apps, not some jewel swapping game.
Good stuff, always like hardware reviews on Anandtech! Feature Request: When you create the charts, can you include the SoC next to the device name? It's good to know that a Google Nexus 10 is faster than Google Nexus 4, but I think the additional information of the SoC contained within would help chart readability greatly. Thanks!
I very much second that, since I was struggling with the very issue, always scrolling up and down in an effort to keep up with which SoC is in which product.
Anyone know the reason why the Galaxy S4 (Exynos 5420) GPU performs slower than the iPhone 5 GPU on offscreen test? Even though the GPU is similar but the S4 having a higher clockspeed and higher bandwidth?
Since performance comes at a cost, this Soc seems in the 8/10 W Tdp range running games. The comparison is pretty clear with A4-5000, done on the same process. Likely the power draw is lower in cpu only tasks but IMO users will enjoy a fast battery death under games. It's a Qualcomm choice, will see the market acceptance. I prefer more balanced socs, this do not seem the case.
I will be interesting to see how the Silvermont based SoC stand up against Snapdragon 800 and Tegra 4. I guess it will do quite well considering that CloverTrail+ seem to beat even Exynos 5 Octa. CT+ is a dual core CPU using a 5 year old Atom design...
I don't feel that today's phone need more CPU-power, they should instead focus on reducing power consumption. But tables can definitely use a lot more CPU-power!!
That link seems to be an Intel sponsored test without any details how it was done. The benchmark I linked however is known to be independent and doesn't favour either Intel or ARM.
How does that link seem Intel sponsored? The numbers presented by the analyst firm is just the numbers from Antutu and that has been verified by others (search YouTube for example).
If you read the review of the Z2460 & Z2670, then notice that Geekbench is really the only benchmark where those SoC performs bad against ARM SoC used at that time (Tegra 3 was very popular at that point).
If you ever actually used units based on Z2460 or Z2670 (tablets) side by side with units using Tegra 3 running the same OS and application, then you would know that the Atom systems definitely feels faster so I don't really trust Geekbench as a measurement on how things feel.
The question one can ask oneself: how come that two benchmarks that are supposed to measure roughly the same attributes of a system get such radically different result? Do you have other benchmarks? The only other benchmarks I can find myself are the web-based ones and they show the Atom in pretty good light.
Having used Apple tablets (with their relatively slow CPU but fast GPUs and hardware accelerated drawing/scrolling), I believe how fast a device feels is more a software/GPU issue than a CPU one. Similarly browsing depends a lot on how fast you can download stuff.
The original ABI press release almost literally reads like an Intel advert. And the chosen benchmark is one of the few where Atom scores well, so that way it not only seems faster then it really is but appears more power efficient as a bonus (as the A15/Kraits have to do more work to get the same score). If you did power numbers on Geekbench then Atom wouldn't look very efficient at all - the only results Intel will show is where they look best.
I have no idea what AnTuTu actually does but Atom scores ridiculously high, so much that either the calculation must be incorrect or there is some kind of cheating going on. For example Atom appears to score more than twice as much as A15 on the memory test, but Geekbench shows a completely different story: Stream - the traditional memory benchmark - shows that A15 is about 3 times faster.
So which is right? Well give me the AnTuTu source code and as a compiler/benchmarking expert I'll explain what is wrong with it and how it could be trivially gamed to show a much higher score (if you look at their web page, it is full of cautions about companies cheating the results). I know for a fact one cannot do this with Stream, which is why I trust the Geekbench results.
I think they want you to pay for it... However just like SPEC they use standard open source benchmarks, so you can find pretty much every one of them if you do a search.
I don't really agree that Antutu is the only benchmark where Atom does very well compared to ARM. Look at AnandTechs review of the Samsung Galaxy S4 and notice how well the Motorola Razr i does against the S4 in a number of benchmarks. Razr i uses Z2460 single core Atom, K900 uses Z2580 dual core Atom with 33% faster RAM-interface and WAY faster GPU.
I do own a iPad4, so I know it is a slick device, but by comparing Intel vs ARM SoC I meant when they are running the very similar software. Compare Z2760 on Win8 vs Tegra3 or Krait on Windows RT or Z2460 on Android vs Tegra3/Krait on Android. Z2760 definitely feels faster and Z2460, despite being single core and used in very cheep phones, feels at least on pair with Samsung S3 performance wise on tasks like webbrowsning and usage of non 3D applications (the GPU suck in Z2460 for 3D).
My guess is that applications like a webbrowser and applications like Antutu put a lot more stress on the L2 cache and the RAM-subsystem and Intel just destroys the competition when it comes the L2 latency and compared to ARM, even Atom has decent bandwidth to L2. Geekbench probably runs very close to 100% out of L1 cache.
The S4 review shows single threaded JavaScript tests, so don't expect core count to make any difference. As I mentioned in another post, Anand shows much lower results for S4 than other sites, so that makes the Razr i seem more competitive.
While Atom and ARM both use Android, Intel does their own builds and modifications for x86. They have a lot of man power so can optimize for the common benchmarks as well as making the user experience better like Apple. I can't comment on your experience but I bet it is not at all related to raw CPU performance.
Geekbench uses the memory system in a serious way, Stream for example measures RAM to RAM copies - if it was L1-L1 copy the scores would be 20 times higher! While Cortex-A9 was behind Atom in L2 latency and some variants had terrible memory performance (ahem Tegra 3), A15 has certainly caught up.
I think it is actually far more likely AnTuTu does some tiny loop that stays entirely within L1 and effectively just measures the frequency of a CPU. They might even multiply single threaded scores by the thread count... Also they seem to be adding up various scores rather than calculating the geometric mean. AnTuTu looks like a broken benchmark.
Geekbench might do RAM-to-RAM, but most programs you run on a mobile will have L2 cache hit-rates way over 90%. According to http://www.7-cpu.com/cpu/Cortex-A15.html, Exynos 5250 has a L2 latency of 21 cycles, Saltwell got 15 cycles and Silvermont 13 cycles latency, don't know if I would call 40% higher latency as "caught up"...
AnTuTu tries to simulate "normal" applications like games and other applications by running what they think is "typical" workloads (you can see 3D-scenes and simulated applications as the benchmark runs), but I cannot say I know anything about the details of how this benchmark actually work.
The problem with most benchmarks i.m.h.o is that they focus WAY to much on performance when all cores run at full tilt. Connect ADB to an Android phone and run "top", do some "normal" tasks and check the load-average and CPU-load from the running processes. You rarely see more than one CPU-core being used, I bet you never see more than two cores used at 100%, so the only thing that matters for how fast the device feels is the integer single threaded performance and even the old in-order Atom is competitive here.
The 64KB access time (= fastest L2 access without TLB misses) went from 23 cycles on A9 to 13 cycles on A15, while Atom N2800 does 11. It's clear A15 has caught up even if Atom still has a slightly lower L2 latency.
Yes performance benchmarks will never give an indication of typical use or how responsive a device is. For example how fast software & hardware ramp up clock frequency from a dormant state to active can make a large difference.
The geekbench seems to me emphasis the kind-of HPC working load. And the AnTuTu is more relative to the mobile usage working load. And the real life mobile user experience is more related to the AnTuTu result than to geekbench result. But anyway, it's really depends on the user case. If some one prefer to use their mobile phone to simulate the nuclear reaction rather than browser the internet or play some game, geekbench is their first choice to judge their phones:) I really admire the develop who gives geekbench it's name:)
And just as a opposite to you, that's why I don't trust the Geekbench result for my mobile choice.
Geekbench does the traditional benchmarks, so in that way it is a bit like a small variant of SPEC. However what does AnTuTu measure? When I go to their website I cannot find any information on it. All I see is endless announcements about people cheating the scores. That really instills confidence...
Given you seem to have a solid trust in AnTuTu, can you explain in more detail what it actually measures? Have you seen the source code? Is it native code, interpreted or JITted? It seems AnTuTu also does memory, integer and FP scores just like Geekbench, so what makes you claim it is more relevant to mobiles?
Why would you expect that? Silvermont is just a 2-way limited OoO core, very similar to Cortex-A9. Cortex-A15 on the other hand is very aggressive with a large reordering window, 3-way and supports 2 memory accesses per cycle (vs just 1 for SilverMont). Given that it should be obvious Silvermont will have a lower IPC than Cortex-A15.
You'll be disappointed. Even Intel is claiming just a 2.8 speedup with double the cores and a 20% clock boost. So at the same core count and frequency the speedup from Silvermont is only 16%...
Qualcomm will probably keep the power low as they are cheating on benchmarks. They are detecting what application is launched. For a given list of benchmarks they are tweaking CPU governor to go to full power, tweak data base access and anything yet uncovered. This way they got very competitive scores on known benchmarks. This is somewhat smart, however other competitor (IC maker) don't do that so get scores more inline with the way the device is configured in order to get a good pert/power compromise.
Some phone makers are also cheating on top of Qualcomm Cheating, at least one.
Note that probably many phone makers using Qualcomm chipset even do not realize they are cheating on behalf of Qualcomm.
In what way is setting the CPU governor cheating? You won't get a better score than the maximum possible score, you'll just get it first time. It seems there is far worse cheating going on with eg. Atoms AnTuTu scores, which are impossibly high (A15 has 3 times faster Stream scores and yet less than half the memory score in AnTuTu?!?).
You're naive if you believe you can accurately estimate the performance of Silvermont based on high level architectural details like the ones you have provided. Given how the current generation of Atom is still using in order execution it should be no match to Qualcomm's Krait or even a basic Cortex A9 (which has at least limited OOO execution) if we were using your logic. Anandtech's own analysis has shown that this is not the case at all.
"Silvermont is just a 2-way limited OoO core, very similar to Cortex-A9. "
And yet even Clovertrail+ crushes the ARM A9 and it is older than Silvermont.
Yes you can make accurate predictions based on the microarchitecture, I've got the experience and patents to make that statement. A9 typically crushes much higher clocked Atoms, eg. http://browser.primatelabs.com/geekbench2/compare/... Both are 4 threads btw, so it is a fair comparison, however a 1.6GHz A9 beats a 2GHz Atom on most single threaded benchmarks as well. Again this is entirely expected based on their respective microarchitectures.
Anand's analysis is not believable given he only ever runs JavaScript benchmarks, and these are highly dependent on the browser used. Even on the same hardware there are huge variations. To give an example, the S4 Octa scores 679 on SunSpider according to FoneArena, but Anand claims 920 above. That's a huge difference on the same phone - who is right? The 679 number is correct as that is close to what the Exynos 5250 scores in the Chromebook. So until Anand cleans up his benchmarking I would take his scores with a bag of salt.
If you think that a single A9 core can beat and destroy a single Atom core, you are wrong.
From Anand's iPad2 review: Alongside multitasking is the performance problem. With the original iPad even deleting several emails at a time was a bit choppy, and web page rendering performance needed tons of work. As always Apple does its best to hide the limitations of the platform but I must point out that even the iPad 2 with a pair of ARM Cortex A9s has lower CPU performance than a netbook with a single core Atom. The fact that you can't really tell most of the time is a testament to Apple's software engineering, but it doesn't change reality.
Some more benchmark: Tegra 250 @ 1.0 GHz Core: 2 Compiler: GCC 4.4.5 Coremark/MHz: 5,148 Total Score: 5148,01 Threads: 4 Flags: -O3 -funroll-loop -falign=8 -mcpu=cortex-a9 -mfpu=vfpv3-d16 -mfloat-abi=soft
Atom N280 @ 1.68 GHz Core: 1 Compiler: GCC 4.4.4 Coremark/MHz: 3,187 Total Score: 5353,79
As you can see, _2_ A9 cores nave total throughput equal to a single Atom core. Atom need high clock of curse, bit it was a specific architectural choice from the start.
The truth is that only A15 decisively defeat Atom.
Unfortunely, Atom GPU was really terrible, being crushed by all others.
Really? Looking at the published results it shows Exynos 4 does 5560 Coremarks/core at 1.4GHz.
The fastest per-core Atom result is 2.3 CM/MHz for 1 thread, and 3.3 with Hyperthreading.
Cortex-A9 does 4.0 for 1 thread - so it is 74% faster single threaded, and 21% faster core for core.
So the A9 destroys Atom on CoreMark as well. I am surprised several of you are trying to argue that in-order cores beat out-of-order cores despite the facts.
No, it is incredible how you pretend to extrapolate _precise_ performance numbers from vague arch details.
Return to Coremark site, because you misunderstan the benchmark results. The CM/MHz score represent the score of the entire soc - so it don't rule out core count differences. Let see the CM/core score instead and you will find that Atom is in the same field of A9 scores, sometime much better.
Some examples: Atom z520 vs Tegra2 and Atom n2800 vs exynos4 quad.
Please also note that: - Coremark does not stress l2/memory in any way. This is the only reason why A9 slow memory interface does not interfere here; - the compiler has enormous importance in it's score.
The real Atom problem was the terrible GPU and companion chipset.
I listed the per core results, as I said A9 is 74% faster single threaded and 21% faster with Hyperthreading enabled. These are results from the EEMBC website, no complex extrapolation involved.
Coremark runs mostly in L1, however it does stress the branch predictor seriously. All benchmarks have a major compiler component. Coremark is horrible like pretty much any EEMBC stuff so I don't think it will become popular.
Atom N2800 @ 1.86 GHz (2 Atom cores) Coremark: 12286.90 Coremark/Core: 6143.45
Atom advantage: 10%
### Note: Why the two A9 and Atom scores are so much different (see Tegra2 vs Exynos and Atom Z530 vs N2800)? The reason lie in the compiler: recent GCC version have greatly improved their efficienty with in-order uarch. Moreover, please also note that the high A9 score (Exynos) was obtained with their specific arm compiler. I am sure that, if benchmarked using Intel C Compiler, the Atom score would be higher.
### Summary: the Atom core is more than capable to compete against A9. You can argue than Atom has an higher clock, but in phone/tablet environmento clocks don't mean nothing. What is important is performance/watt. This bring us to the two real Atom's problem: 1) a very low efficiency chipset and low integration. Moorestown (intel first attempt to mobile with Atom) was doomed from the start because it require 4/5 chips to enable a full-featured phone;
2) a very slow GPU (with very bad performance/watt).
Moreover, it is widely understand that A9 OoO engine is a mild implementation only. A15 is much stronger in this reguard, sometime (not too often, anyway) even apporaching AMD Bobcat single-thread performance.
No - the performance comparisons that are useful are:
1. Max score for a SoC - despite running at a far lower clock, in both comparisons A9-based SoCs win by more than 80% in overall performance. 2. Efficiency of a core at the same frequency (IPC) - Without Hyperthreading A9 is 74% faster, with Hyperthreading A9 wins by more than 20%.
Note that your comparison doesn't work. You can't come to a conclusion about A9 vs Atom performance when you compare with wildly different frequencies. Also it means giving Atom the advantage of having 2 threads vs 1 on A9. So to make the comparison fair you need to compare with an equal number of threads or at the same clock.
Yes GCC has improved a lot in recent years, on ARM it has become a reasonable compiler and competitive with ARM's armcc compiler. I don't know how much better ICC would be on Atom, but I suspect the gap is far smaller as well.
A9 is not hugely OoO indeed, just like Silvermont. A15 is aggressive OoO and beats Jaguar.
No, I don't agree again. You explicitly talket about CortexA9 and Atom uarch, _not_ their SoC implementation.
You can not use the total SoC score as uarch benchmark - simply because it don't rule out differences in cores number. To measure uarch performances you need to do a core-by-core comparison. Let me do an example: using total SoC score, a 4xA9 SoC is faster then 2xA15 one. However, the latter uarch is considerably more advanced.
A very similar argument can be done for frequency: Atom was _from the start_ designed to hit a relatively high-clock, yet low power target. This was deliberately done to exploit Intel 45/32nm HKMG process, which don't scale power down much for lower frequency target. It is simply a question of design targets: for low power chips, you can get (relatively) high-freq _or_ (relatively) high IPC - not both (actually).
So, you must decide: are you comparing uarch of final SoC implementation? Because, from an uarch point, Atom win. From a performance/watt metric, their bare cores tend to be on par. From a final product specification, A9 is way better because there are many high-integrated, low power, low cost SoCs from a multitude of vendors. On contrast, Atom-based SoCs are offered only by Intel and with a much lower integration factor (and higher cost) - until now,where they latest platform begin to be very competitive against older A9 SoC.
The "little problem" is that ARM is shipping with 2x and 4x A15 cores, and against them Atom is a disvantage.
While Atom was indeed designed for high frequency, A9 reaches higher frequencies: Atom maxes out at 2GHz on 32nm, while A9 does 1.7GHz on 40nm and 2.3GHz on 28nm. So you can't claim a "microarchitecture" win for Atom when you compare against a low clocked A9.
Secondly, since you argue that frequency is an important aspect of the microarchitecture, I would argue that core count matters equally. A9 was designed to be simple and small, so it is typically used as a quad-core. On the other hand Atom is a large and complex core which uses Hyperthreading rather than multiple cores. So if you want to do a fair comparison with Hyperthreading enabled then you have to use 2 A9 cores for every Atom core. That's how they have been designed to be used.
What is the difference between a module, a HT enabled core and a dual core? These are just different ways of improving multithreaded performance with different hardware tradeoffs - but to software they all appear identical.
In conclusion: you cannot just pick whatever comparison you want. Either you compare the whole SoC, including its frequency as well as core count, or you compare microarchitectures normalized on core count and frequency. You can't include one but not the other as frequency, core count and TDP are related.
So, you started about in-order vs OoO and now you are speaking of die size and perm/mm2?
1) While CortexA9 was rated for 2 GHz operation, a single A9 core would dissipate more than 2 Watt at this frequency. Atom is not so much different in this reguard. Moreover, can you point me a phone that use a 2 GHz A9 implementation? I bet no.
2) Atom is also MP form the start: it has the same bus unit and MP capability of Netburst uarch. By which metrics these are inferior to the ARM MP implementation?
3) By die size comparison, A9 is clearly better then Atom. However, its performance are lower.
4) HT is simply a smart sharing of some key structure in order to interleave two thread on the same core. You can not count HT as another core. For example, barrel microprocessors can interleave many threads on a single core: Sun T1 can inteleave 4x threads per core, T2 8x core. Do you count T1 as having 32 cores? If so, you are wrong.
Both I and other users pointed you many reviews and benchmarks where Atom is clearly identified as faster then A9. However, you contine to change metrics.
The only benchmark that paint a different picture is Geekbench, which show A9 in the same league as Sandy Bridge. Do you _really_ think this is true? In SPEC benchmarks, SB is quite close to the big, power hungry but powerfull POWER7. Do you really think that A9 is remotely comparable to this core? Really?
I already stated this: if you compare SoCs, well, A9 wins, because there are many well done SoCs based around it. However, from uarch/performance side, Atom wins.
The funny thing is that is now totally irrelevant: A9 is superseeded by A15, and Atom is very near its EOL. Moreover, Jaguar seems to be a very competent table chip.
Unfair to compare the A9's to Atom. The Tegra 2 was a old revision of A9 while lacking NEON etc. The newer A9 are more fair to compare. Also a single A9 at 2Ghz wont produce 2 watts at all, the 2.3Ghz Tegra 4i would be worse than the A15 if it did. Remember the nm is 28 not the old 40's.
Coremark can be vectorized, which one reason why the scores have improved. Tegra 2 was terrible in most respects, Exynos 4 and Calxeda are currently the best A9 implementations until Tegra 4i is released. Obviously it will be very interesting to see how Silvermont will do against Tegra 4i.
Now we've agreed that A9 can reach similar frequencies then we can normalize on frequency, yes? When you normalize on frequency then in-order vs OoO matters more, which is what I said all along.
1. NVidia has shown their 2.3GHz Tegra 4i reference phone which should be on the market soon. The reason you don't see 2GHz A9 phones today is (a) use of older processes and (b) use of quad cores which leads to too much power use at high frequencies. If they were single or dual core then it would be easier to go for 2GHz.
2. Atom only supports dual core (even Centerton is dual core trying to compete with the Calxeda quad core...), and until the recently released K900 I believe all Atom phones were single core. On the other hand most smart phones have been quad-core since 2012.
3. I still don't see where you think that A9 is slower per core than Atom. At the same frequency it is unbeaten even if you give Atom the advantage of Hyperthreading.
4. If you say that a core with HT is just a single core then surely it should look like a single core to software as well and provide a speedup to single-threaded applications? As soon as you enable HT for a multithreaded application then it would be unfair to compare with single threaded performance on a non-HT core. You can't just pick and choose a comparison that gives one core a major advantage over another.
So far I haven't seen any evidence of native benchmarks where Atom is clearly faster. If you do a fair like for like comparison then A9 wins every time. However if you remove the advantages from one core but keep all the advantages for another then you can of course prove whatever you like.
1) During Atom lifetime as competent product (read: before A15 release), _no_ phone (or tables, to best of my knowledge) shipped with A9 faster then 1/1.2 GHz. Even Scorpion based phones had higher frequencies. Anyway, from the Coremark comparison written above, you can see that normalizing core/frequency lead to similar per-core performance between A9 and Atom. With the difference that Atom shipped at generally higher frequency, even @ 2.0 GHz: http://www.anandtech.com/show/6307/motorola-hits-2...
Comparing a 5-year-old Atom to a to-be-shipped Tegra 4i which as tweaked cores seems fair? I think no.
2) For scaling above 2 cores, currently Atom use FSB communication. However, I can not see why you need 4 Atom cores in a phone. Anyway, you have a point here (on-chip communication are much more pratical) ;)
3) HT is a simple, cheap method to increase total tput. You can not consider an HT-enabled Atom as having 2 cores. Even on coremark sites an HT-enabled Atom is considered a one-core processor. I reiterate: how to consider a barrel processor? A many core approach?
You can not mark all web browser benchmark as "flawed". Sure they are not 100% uarch dependant, but they are a good method to evaluate the entire software stack used by these processors.
1. Not true - A9 was running at 1.4GHz in 2011 products, 1.5GHz early 2012 and 1.7GHz late 2012. So A9 was most definitely available at higher frequencies, including in phones (One X). With the shrink to 28nm it gets another boost to 2.3GHz. So in frequency it can match or beat Atom (note by the time it comes out, it'll be compared with Silvermont which achieves similar frequencies).
3. Yes a barrel processor looks and behaves like multiple slow cores. The fact that they share hardware in a very specific way is an implementation detail. An equivalent approach would be multiple independent cores. Do you consider a Bulldozer module as a single core as well?
If you prefer to just compare a single core then you should only consider its single threaded performance. As soon as you compare multithreaded performance, then it is completely fair to allow each CPU designer to decide how to implement that, whether it is with HT/SMT, multiple cores or AMD's inbetween approach. You're right of course that HT is a fairly cheap way of adding more threads if you already have a complex core - using multiple cores is only a good idea if you designed your core to be as small as possible.
4. Anand is the worst place to look for CPU benchmarks. Most are JavaScript benchmarks which can vary by a factor of 2 even on the same browser. And his Linpack test is more a Dalvik test than about FP performance. Although there are cases where performance of a browser is important (things like rendering pages), we are discussing micro architecture performance, not how the Android software stack performs.
1) Almost all widespread phone had A9 running at 1.0/1.2 Ghz for so much time. Only recently they increased their frequency at about 1.4 Ghz, and atom is @ 2.0 GHz now.
2) "Look as" is not the same as "it is". So a T2 is the equivalent of a 128 core machine? Really? You are confusing SMP with SMT. Look at coremark results: a quad core A9 has performance 4x of a single core, and the score/core reflect this (it is the total score divided by 4). Quad thread Atom instead is a two core uarch, only slightly (30%) faster when tested with 4 thread instead of 2. In fact, coremark/core score is obtained dividing the total score by two, even if the test run with 4 threads. Even coremark developers are wrong?
3) All the posted benchmark,with the exception of Geekbench, show how Atom single core performance are better then A9. Yet you refuse to look at the data, dismissing them as "flawed" or inappropriate. Even user experience, in your eyes, does not matter. can you point us to some reliable bench?
I think Wilco gave up on discussing with you since you just proved your huge ignorance (the A9 clock speed) and are even trying to defend it. When was the One X + released? What about the Tansformer prime infinity?
I think Wilco gave up on discussing with you since you just proved your huge ignorance (the A9 clock speed) and are even trying to defend it. When was the One X + released? What about the Tansformer prime infinity?
Great performance. Now if only we could get some more software optimizations on Android... My iPhone 5 iOS 7 beta 3 results: Sunspider 1.0 - 709.0ms; Kraken - 13783.9ms; Octane v1 - 3056; Browsermark 2.0 - 3056. So 9 month old dual core hardware that's faster than anything on the market today, and faster in some benchmarks than something that's not even on the market yet... Google, wasssssssup??
Thanks Brian for a great test of this chipset. I am especially interested in the video encoding performance, and your inclusion of the 'MediaInfo' screen capture is really useful to see how it is encoding H.264 video.
Your MediaInfo clip clearly shows that this chipset can encode H.264/AVC at 2160p @ 25fps @ 120Mbps, (Baseline @Level 5.1). As a low-budget film-maker, I am speculating about the possibility that the encoder could alternatively be configured to handle 1080p @ 30fps (perhaps 60fps) @ 4:2:2 colour sampling @ 10-bit (perhaps 12-bit) depth. I have not been able to get confirmation of this, but if so - at this price point, this chipset could potentially unlock high quality video capture on regular consumer-level DSLR-type cameras; something that has been limited to commercial broadcast cameras (at high-budget prices) up to now. If anyone is familiar enough with AVC profiles and Levels (and related matters) to be able to speculate about this, I would like to hear your thoughts. Thanks again to Brian.
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.
115 Comments
Back to Article
akbisw - Tuesday, June 18, 2013 - link
YESSS!!! THANK YOU QUALCOMM exactly What I expected!!!Piano Man - Tuesday, June 18, 2013 - link
Nice performance, but the lack of power utilization benchmarks prevents me from breaking out the party favors for the time being.clemsyn - Tuesday, June 18, 2013 - link
I agree, power utilization and heat are very important in mobile. Hope this don't come with a fan :)Xilliox - Wednesday, June 19, 2013 - link
I have been working on this for a long time, trust me, it does not need a fan!FwFred - Tuesday, June 18, 2013 - link
Comparing phones to a tablet is a bit odd. Glad there was a Nexus 10 at least.kcsween - Wednesday, June 19, 2013 - link
You all almost sound like this is Qualcomm's first go at making a mobile processor. I'd find it very hard to believe that they'd not take into consideration power consumption and heat or any other negative factors they considered when they made their other chips. Having said that, all that could be out the door in the race bigger and better, who knows. Guess we all will have to wait and see how this things performs in an actual smartphone.arkhamasylum87 - Tuesday, June 18, 2013 - link
How does this compare to the Ivybridge or Haswell Y series processors in terms of CPU/GPU performance? Would like to see those comparisons. So we can extrapolate how this will compare to a fanless Core product from Intel. The gap is there in terms of connectivity but the perf gaps on CPU/GPU would be interesting to note.teiglin - Tuesday, June 18, 2013 - link
Ivy Y no longer seems interesting to me but for the moment, Haswell 15W (both HD4600/5000) would be super interesting to add to the "great equalizer" page of charts. Doesn't Anand have an Acer with i7-4500U as well as the MBA with i5-4250U? Demand he run some benchmarks for you so you can update those charts. The comparison to i5-3317U made sense against APQ8064T, but 8974's contemporary is Haswell, not Ivy.And when you have Haswell Y samples, those would be neat, too. :)
jeffkro - Wednesday, June 19, 2013 - link
15W still requires a fan.sna1970 - Tuesday, July 9, 2013 - link
not really,the Haswell can run from 6 watts up to 15 watts.
Glindon - Tuesday, June 18, 2013 - link
I wonder how Rouge is going to compare to this. My guess is Apple will ship a rouge powered iPhone and iPad before devices start shipping with Qualcomm.xTRICKYxx - Wednesday, June 19, 2013 - link
Do you mean Rogue?Glindon - Wednesday, June 19, 2013 - link
Doh! Yeah, Rogue.MrCommunistGen - Tuesday, June 18, 2013 - link
*Drool* I also love the level of transparency we're getting from Qualcomm even compared to a product cycle or two ago.bleh0 - Tuesday, June 18, 2013 - link
From what I'm seeing here this an amazing chip overall. That GPU really surprised me but I'm gonna wait till the battery life/power draw can be tested.nunomoreira10 - Tuesday, June 18, 2013 - link
if this thing keeps it´s power restrained to 3-4 watts the gpu perfomance per watts is astonishing.Alketi - Tuesday, June 18, 2013 - link
Wow, crushing performance. Two things stand-out.1. It's beating it's almost doubling it's closest Android competitor on some benchmarks.
2. A 1-year old Apple chip is still killing some benchmarks.
Way more exciting than the ~15% x86 gains / year.
One thing's for sure, the upcoming ARM-based HTPC's are going to absolutely kick ass.
jeffkro - Wednesday, June 19, 2013 - link
The hardware is getting there for ARM HTPC but we need an Android OS with DVR functionality.karasaj - Tuesday, June 18, 2013 - link
And this is why I'm dying to get an internship with Qualcomm next summer... they're going to become the Intel of mobile :)tuxRoller - Tuesday, June 18, 2013 - link
They already are. Look at they're market cap.althaz - Tuesday, June 18, 2013 - link
*Their*I'm not sure you could call them the Intel of Mobile. There's a couple of key differences - the first and most obvious being that Intel is leading in x86 IPC, power management and, this is the big difference, manufacturing. Intel won't and can't go away because they have a tangible advantage over their competitors that will require incredible investment to overcome.
The other is that Qualcomm have a large slice of an ever-increasing pie, but Intel have basically the whole x86 cake to themselves, with the crumbs falling to AMD.
It's probably also worth mentioning that Qualcomm seem better at producing the right improvements for consumers, wheras Intel were late to the low-power party and slow to get going on graphics to (although both are clear foci of the company now).
Interestingly, I believe Qualcomm's mobile division is basically AMD/ATI's old mobile division, which they didn't think was worth it to keep. That's looking a like a pretty huge mistake right now.
tuxRoller - Wednesday, June 19, 2013 - link
Thanks for the grammar correction, teach. Now that I know it any way you can teach it to my phones keyboard? Grammar is useful for helping to avoid ambiguity but the previous could only be understood one way. IOW, your correction wasn't needed unless you were just being anal.You make some good points and it's clear that not clear what it means when someone says "X is the Intel of Y". Intel using their fab is certainly a part of their success. That they are THE key x86 player is, I think, the more important touchstone. It was that which I was speaking of when I spoke of them. I THINK they are, by far, the biggest, most profitable supplier of ARM based SoC in the world. Samsung is huge but they consume what they make. TI is leaving the mobile field and haven't had a major design win for awhile.
BTW you're right about ATI and adreno but the 300+ series is vastly different from the architectures Qualcomm inherited from ATI.
FwFred - Tuesday, June 18, 2013 - link
Very inconclusive without any power measurements. In its tablet reference form factor certainly beats modern SoCs in phone form factors. I'd like to see it against Tegra 4, Bay Trail, and Haswell (4.5W SDP) in a tablet form factor.grahaman27 - Tuesday, June 18, 2013 - link
I second that.aryonoco - Wednesday, June 19, 2013 - link
I agree, I want to see it next to Tegra 4, Bay Trail, and also Kabini and Temash.But I have to say, it looks good, really good. Especially that GPU.
To think that Adreno was AMD/ATI's embedded graphics division that AMD didn't think was worth anything and sold for peanuts (something like $30M IIRC) to Qualcomm. Dirk Meyer really ruined that company.
FwFred - Wednesday, June 19, 2013 - link
Doh, totally forgot about Temash. I was trying to think of all the fanless tablet SoCs coming out.wsw1982 - Thursday, June 20, 2013 - link
The engadget has the benchmark for snapdragon on phone and tablet. It can be seen that the snapdragon 800 is very fast in tablet setup, but only little bit faster than snapdragon 600 in phone setup. I guess this partly answered your TDP concern.http://www.engadget.com/2013/06/18/qualcomm-snapdr...
iwod - Tuesday, June 18, 2013 - link
I Love the Equalizer Part. Finally we have some sense of how far we are from Mainstream Desktop Performance. Would be great if Haswell were added to the Chart, and a UL Haswell running at similar Power usage in there as well. ( Although i believe Intel would properly pressure Anand not to have done so )Impulses - Wednesday, June 19, 2013 - link
Can't believe how far behind Atom's GPU is at this point and how long Intel sandbagged the market with it...I've been rather disconnected from the Android news/rumor mill, any chance Snapdragon makes it into the next small Nexus tablet design? I really wanna downsize/upgrade from my OG Transformer...
Impulses - Wednesday, June 19, 2013 - link
Meant Snapdragon 800, obviously.Bob Todd - Wednesday, June 19, 2013 - link
That would be awesome, but I highly doubt it. Unless they radically change their target market for that device, I don't think you'll see a bleeding edge SOC or super high resolution display. It was meant to win back the market share lost to the likes of the budget friendly Kindle/Nook tablets who were eating Google's lunch in the tablet market while using their OS at the core. I just don't see a Snapdragon 800 in a ~$200 device, for the same reasons you don't see Swift or Retina displays in the iPad Mini. I think we're much more likely to see an S4 Pro quad/Adreno 320 (possibly Snapdragon 600?). Even that would be a hell of an upgrade to the existing Tegra 3, and I'd happily throw my Nexus 7 up on eBay for that. Admittedly, I may end up with an x86 Windows 8.1 tablet with Silvermont or Temesh instead of the next Nexus 7 anyway.Krysto - Wednesday, June 19, 2013 - link
"Snapdragon 800 can definitely be quicker than ARM's Cortex A15"What does that sentence even mean? You're comparing 2.3 Ghz Krait 400 vs 1.7 Ghz Cortex A15.
aryonoco - Wednesday, June 19, 2013 - link
He's comparing what he has (Snapdragon development tablet) with other things that he has.Clock frequency is pointless to compare by itself. Sure A15 might be faster clock-for-clock, but it's also more power hungry so you won't be seeing it hit 2.0 Ghz anytime soon. Looks like a Snapdragon 800 can easily clock higher which more than compensates its IPC disadvantage compared to A15 while keeping power usage in check.
Qualcomm is sitting very comfortable right now. No wonder their market cap is now bigger than Intel, while they have a fraction of Intel's revenue/profits. The market clearly likes what they see from them.
Wilco1 - Wednesday, June 19, 2013 - link
A15 is already at 1.9GHz with Tegra 4. If A15 was built on 28nm HPM then it would also reach 2.3-2.5GHz. I bet we'll get something like that this year.Note there is no evidence A15 uses more power. In S4 tests it seems the Exynos Octa has lower power consumption than Krait.
Wilco1 - Wednesday, June 19, 2013 - link
Just came across this: http://arstechnica.com/gadgets/2013/06/qualcomms-s...Looks like Tegra 4 still beats it at 1.9GHz. And good to see some tech sites showing native benchmarks.
xinthius - Wednesday, June 19, 2013 - link
That isn't the impression I got from the article. CPU wise Tegra 4 edges it, but on the off-screen GPU tests Adreno has a solid lead.Wilco1 - Wednesday, June 19, 2013 - link
Tegra 4 finally has a fast GPU - nobody has benchmarked it yet but the one available score (Egypt HD at 57fps) suggests it will be a close match with Adreno.sherlockwing - Wednesday, June 19, 2013 - link
Tegra 4 is already A15 on HPM and it is capped at 1.9Ghz, no evidence on how much further it can be pushed. Also Exynos Octa have 4 A7s to reduce power consumption, in GSMArena's benches it still fall behind S600 in battery life.Wilco1 - Wednesday, June 19, 2013 - link
Anand says it is LPM (unlike Tegra 4i, which is HPM): http://www.anandtech.com/show/6550/more-details-on...The Exynos variant of S4 does indeed fall behind in battery life in the GSMArena test, but that was on talk time and video playback, ie. not related at all to the CPU. QC has a big lead in low-power modems and it shows. Browsing was within 6%, so that confirms that the A15 uses similar power as Krait to get slightly better performance. I'm still hoping Anand will do a full review...
aryonoco - Wednesday, June 19, 2013 - link
Yes, the architecture comparison between A15 and Krait 400 and next-gen Swift will be very interesting.Add in Bay Trail and Temash and we CPU enthusiasts are really experiencing a golden age. When was the last time we had so many different competing designs and architectures? When you compare the progress from Scorpion to Krait 200 to Krait 400 (or A9 to A15) in a matter of 2 years... the pace of development in the mobile world is astonishing.
I'm also relying on Anand to do some full analysis and review of all these architectures once they become available. It's pretty clear that Tegra 4 is going to be cheaper than Snapdragon 800, it would be great to see them side by side, with power analysis and all.
Krysto - Wednesday, June 19, 2013 - link
"The overall graphics score from Adreno is amazing. We're looking at almost 2x the next fastest contender here, the Adreno 320 based Snapdragon 600."Why is that so amazing? Adreno 320 came out first last fall, which is about a year before S800 with Adreno 330 comes out. Are you saying you're surprised mobile GPU's can double in performance every 12 months?
nunomoreira10 - Wednesday, June 19, 2013 - link
On the same process nod, hell yeahAenean144 - Wednesday, June 19, 2013 - link
Not exactly the same process node, and we don't know TDP yet.richworks - Tuesday, June 25, 2013 - link
please excuse my ignorance but how is TDP related to the process node?dishayu - Wednesday, June 19, 2013 - link
In the great equalizer charts, can someone please explain how GeForce 8500 manages to outperform the 7900 GTX in DXBenchmark 2.7? The 8500 had similar performance to a 7600GS from what i remember.nunomoreira10 - Wednesday, June 19, 2013 - link
Uniform shaders i believeRyan Smith - Wednesday, June 19, 2013 - link
Unified rather than uniform, but yes. We did an article in April looking at the issue in greater depth: http://www.anandtech.com/show/6877/the-great-equal...Speedfriend - Wednesday, June 19, 2013 - link
Where will the new generation Atom show up in the list? As I understand it, it will have graphics comparable to the HD4000 in the Ivy Bridge, so in theory it should outperform this Qualcomm chip?psychobriggsy - Wednesday, June 19, 2013 - link
No, it will have graphics based upon the HD4000 architecture - but it will have fewer units and run at a lower speed, in order to attain TDP requirements.FwFred - Wednesday, June 19, 2013 - link
The number of EUs does not directly tie to TDP. See HD5000 vs HD5100. The clock speed is important, and wider but slower is more efficient.xaml - Wednesday, June 19, 2013 - link
Thanks for sneakily including a comparison of the two Galaxy S4s as well.etre - Wednesday, June 19, 2013 - link
It seems to me that everything is depending on the power budget, as the dual core A15 inside Nexus 10 is faster in some tests then the quad A15 in S4. Or I am wrong ?mczak - Wednesday, June 19, 2013 - link
Yes in some tests (like 3dmark physics) it is very obvious the quad A15 aren't running anywhere close to their max frequency. This is not surprising last time I looked at some power draw figures just two A15 cores at max frequency alone (so without graphics) could already exceed the power budget of the whole chip.So we'll see how that new Snapdragon chip compares once it's inside a smartphone. Should still be plenty fast but some scores might suffer (maybe we're going again to see freezer vs. no freezer results :-)).
sherlockwing - Wednesday, June 19, 2013 - link
The answer is simple, Dual Core A15s are running on Tablet TDP limits(4W) while the Quad core A15 in Exynos Octa are running on Phone TDP limits(<2W).HisDivineOrder - Wednesday, June 19, 2013 - link
These SOC GPU's are rapidly catching up to PC GPU's. They aren't there yet, but they've made impressive strides in just a few short years.Consoles better start more rapid update cycles or they are going to be left behind at this rate.
sireangelus - Wednesday, June 19, 2013 - link
I just wanted to make people realize something:Assassin's creed 1 & 2 run very well on an old x1950pro, and this chip seems to me that it's faster. When they said that soon mobile platform would exceed the(actual but) old consolle generation they weren't kidding.douglord - Wednesday, June 19, 2013 - link
This looks great - but frankly I need to see this vs Tegra 4 and A7. Also - I agree we need to see vs HD5000/HD5200 Haswell.Tegra 4 is probably a toss up performance wise - and will come down to power (which Qualcom normally wins). Kinda makes me sorry I pre-ordered a Shield when this will be in the Note 3 and I could just attach a controller. Hopefully Nvidia can do a better job of getting AAA games to actually support Shield. I'm sick of all the blockbusters being iOS only (or 6 months early).
Apple seems to double performance every year, and with PowerVR 6 - you could get a bigger bump. That would easily surpass this. If AAPL would just come out with a larger screen, I'd go back to iOS this time around.
And IF we see Haswell tablets with 7-9 hours of battery life - I'm not sure all of this stuff doesn't get relegated to the dust bin of history with Intel taking 90%+ of the tablet/hybrid market.
Anyone have any info on Saltwell's GPU? I love my Latitude 10's ability to run full Windows programs - particularly Office. The CPU beats anything in this article and the power use is low enough I get 16 hours of battery life. But the GPU is so far behind all of these other chips even a 2x-3x improvement isn't going to make it competitive. Why can't they throw in whatever Apple is going to use in the A7. They own part of PowerVR too!
aicom - Wednesday, June 19, 2013 - link
I'll be interested to see how the actively-cooled Tegra 4 in Shield performs compared to the standard passively-cooled stuff. I'd suspect that if you're going to be gaming for a while, the Shield may pull away. Tablets and smartphones these days aren't designed to run at full tilt all the time.The most interesting thing about this generation of SoCs to me is the CPU performance. I don't game all that often, so I'd much prefer a device that has CPU performance to spare. It's the opposite of the problem that we had on the desktop years ago. Now we have GPU performance sufficient for UI acceleration, media decoding, and mainstream gaming, but we still have CPU performance that holds the platform back.
krumme - Wednesday, June 19, 2013 - link
No your Atom does not beat anything worthwhile in this world and especially not the s800:http://browser.primatelabs.com/geekbench2/compare/...
"Intel taking 90% of the tablet/hybrid market"
Ahemm. And how exactly are they going to do that. With a usd 150 APU sans LTE/3G integrated?
Speedfriend - Thursday, June 20, 2013 - link
No with a processor offering 4x the performance of anything that ARM offers. Believe it or not, most corporate users want something that can run enterprise level apps, not some jewel swapping game.Wilco1 - Thursday, June 20, 2013 - link
A low clocked single or dual Haswell won't be anywhere close to 4 times faster than a high clocked quad A15/Krait.jeffkro - Wednesday, June 19, 2013 - link
Cool, should make a great chromebook processor.WagonWheelsRX8 - Wednesday, June 19, 2013 - link
Good stuff, always like hardware reviews on Anandtech!Feature Request: When you create the charts, can you include the SoC next to the device name? It's good to know that a Google Nexus 10 is faster than Google Nexus 4, but I think the additional information of the SoC contained within would help chart readability greatly.
Thanks!
Kill16by9TN - Thursday, June 20, 2013 - link
I very much second that, since I was struggling with the very issue, always scrolling up and down in an effort to keep up with which SoC is in which product.MrPhilo - Wednesday, June 19, 2013 - link
Anyone know the reason why the Galaxy S4 (Exynos 5420) GPU performs slower than the iPhone 5 GPU on offscreen test? Even though the GPU is similar but the S4 having a higher clockspeed and higher bandwidth?Cotita - Wednesday, June 19, 2013 - link
I don't think Tegra 4 will be able to compete here. Tegra 5+kepler will be its last chance.Alberto - Wednesday, June 19, 2013 - link
Since performance comes at a cost, this Soc seems in the 8/10 W Tdp range running games. The comparison is pretty clear with A4-5000, done on the same process.Likely the power draw is lower in cpu only tasks but IMO users will enjoy a fast battery death under games. It's a Qualcomm choice, will see the market acceptance.
I prefer more balanced socs, this do not seem the case.
virtual void - Wednesday, June 19, 2013 - link
I will be interesting to see how the Silvermont based SoC stand up against Snapdragon 800 and Tegra 4. I guess it will do quite well considering that CloverTrail+ seem to beat even Exynos 5 Octa. CT+ is a dual core CPU using a 5 year old Atom design...And this is the most interesting part "Intel's chip scores highest on the most tests and draws the least current"
http://www.androidauthority.com/intel-clovertrail-...
I don't feel that today's phone need more CPU-power, they should instead focus on reducing power consumption. But tables can definitely use a lot more CPU-power!!
Wilco1 - Wednesday, June 19, 2013 - link
Exynos Octa beats the current fastest Atoms by a huge margin at a lower clock speed, so don't expect Silvermont to come close to A15: http://browser.primatelabs.com/geekbench2/compare/...That link seems to be an Intel sponsored test without any details how it was done. The benchmark I linked however is known to be independent and doesn't favour either Intel or ARM.
virtual void - Wednesday, June 19, 2013 - link
How does that link seem Intel sponsored? The numbers presented by the analyst firm is just the numbers from Antutu and that has been verified by others (search YouTube for example).If you read the review of the Z2460 & Z2670, then notice that Geekbench is really the only benchmark where those SoC performs bad against ARM SoC used at that time (Tegra 3 was very popular at that point).
If you ever actually used units based on Z2460 or Z2670 (tablets) side by side with units using Tegra 3 running the same OS and application, then you would know that the Atom systems definitely feels faster so I don't really trust Geekbench as a measurement on how things feel.
Cortex A15 is not that far behind Core i5-3320M according to Geekbench, which to anyone that used systems using those CPUs know is kind of ridiculous.
http://browser.primatelabs.com/geekbench2/compare/...
The question one can ask oneself: how come that two benchmarks that are supposed to measure roughly the same attributes of a system get such radically different result? Do you have other benchmarks? The only other benchmarks I can find myself are the web-based ones and they show the Atom in pretty good light.
Wilco1 - Wednesday, June 19, 2013 - link
Having used Apple tablets (with their relatively slow CPU but fast GPUs and hardware accelerated drawing/scrolling), I believe how fast a device feels is more a software/GPU issue than a CPU one. Similarly browsing depends a lot on how fast you can download stuff.The original ABI press release almost literally reads like an Intel advert. And the chosen benchmark is one of the few where Atom scores well, so that way it not only seems faster then it really is but appears more power efficient as a bonus (as the A15/Kraits have to do more work to get the same score). If you did power numbers on Geekbench then Atom wouldn't look very efficient at all - the only results Intel will show is where they look best.
The Phoronix benchmarks show basically the same as Geekbench: http://www.phoronix.com/scan.php?page=article&...
I have no idea what AnTuTu actually does but Atom scores ridiculously high, so much that either the calculation must be incorrect or there is some kind of cheating going on. For example Atom appears to score more than twice as much as A15 on the memory test, but Geekbench shows a completely different story: Stream - the traditional memory benchmark - shows that A15 is about 3 times faster.
So which is right? Well give me the AnTuTu source code and as a compiler/benchmarking expert I'll explain what is wrong with it and how it could be trivially gamed to show a much higher score (if you look at their web page, it is full of cautions about companies cheating the results). I know for a fact one cannot do this with Stream, which is why I trust the Geekbench results.
tuxRoller - Wednesday, June 19, 2013 - link
Where is the code for geekbench? I've been to their site but was unable to find any links.Wilco1 - Thursday, June 20, 2013 - link
I think they want you to pay for it... However just like SPEC they use standard open source benchmarks, so you can find pretty much every one of them if you do a search.virtual void - Thursday, June 20, 2013 - link
I don't really agree that Antutu is the only benchmark where Atom does very well compared to ARM. Look at AnandTechs review of the Samsung Galaxy S4 and notice how well the Motorola Razr i does against the S4 in a number of benchmarks. Razr i uses Z2460 single core Atom, K900 uses Z2580 dual core Atom with 33% faster RAM-interface and WAY faster GPU.I do own a iPad4, so I know it is a slick device, but by comparing Intel vs ARM SoC I meant when they are running the very similar software. Compare Z2760 on Win8 vs Tegra3 or Krait on Windows RT or Z2460 on Android vs Tegra3/Krait on Android. Z2760 definitely feels faster and Z2460, despite being single core and used in very cheep phones, feels at least on pair with Samsung S3 performance wise on tasks like webbrowsning and usage of non 3D applications (the GPU suck in Z2460 for 3D).
My guess is that applications like a webbrowser and applications like Antutu put a lot more stress on the L2 cache and the RAM-subsystem and Intel just destroys the competition when it comes the L2 latency and compared to ARM, even Atom has decent bandwidth to L2. Geekbench probably runs very close to 100% out of L1 cache.
Wilco1 - Thursday, June 20, 2013 - link
The S4 review shows single threaded JavaScript tests, so don't expect core count to make any difference. As I mentioned in another post, Anand shows much lower results for S4 than other sites, so that makes the Razr i seem more competitive.While Atom and ARM both use Android, Intel does their own builds and modifications for x86. They have a lot of man power so can optimize for the common benchmarks as well as making the user experience better like Apple. I can't comment on your experience but I bet it is not at all related to raw CPU performance.
Geekbench uses the memory system in a serious way, Stream for example measures RAM to RAM copies - if it was L1-L1 copy the scores would be 20 times higher! While Cortex-A9 was behind Atom in L2 latency and some variants had terrible memory performance (ahem Tegra 3), A15 has certainly caught up.
I think it is actually far more likely AnTuTu does some tiny loop that stays entirely within L1 and effectively just measures the frequency of a CPU. They might even multiply single threaded scores by the thread count... Also they seem to be adding up various scores rather than calculating the geometric mean. AnTuTu looks like a broken benchmark.
virtual void - Thursday, June 20, 2013 - link
Geekbench might do RAM-to-RAM, but most programs you run on a mobile will have L2 cache hit-rates way over 90%. According to http://www.7-cpu.com/cpu/Cortex-A15.html, Exynos 5250 has a L2 latency of 21 cycles, Saltwell got 15 cycles and Silvermont 13 cycles latency, don't know if I would call 40% higher latency as "caught up"...AnTuTu tries to simulate "normal" applications like games and other applications by running what they think is "typical" workloads (you can see 3D-scenes and simulated applications as the benchmark runs), but I cannot say I know anything about the details of how this benchmark actually work.
The problem with most benchmarks i.m.h.o is that they focus WAY to much on performance when all cores run at full tilt. Connect ADB to an Android phone and run "top", do some "normal" tasks and check the load-average and CPU-load from the running processes. You rarely see more than one CPU-core being used, I bet you never see more than two cores used at 100%, so the only thing that matters for how fast the device feels is the integer single threaded performance and even the old in-order Atom is competitive here.
Wilco1 - Friday, June 21, 2013 - link
The 64KB access time (= fastest L2 access without TLB misses) went from 23 cycles on A9 to 13 cycles on A15, while Atom N2800 does 11. It's clear A15 has caught up even if Atom still has a slightly lower L2 latency.Yes performance benchmarks will never give an indication of typical use or how responsive a device is. For example how fast software & hardware ramp up clock frequency from a dormant state to active can make a large difference.
wsw1982 - Thursday, June 20, 2013 - link
The geekbench seems to me emphasis the kind-of HPC working load. And the AnTuTu is more relative to the mobile usage working load. And the real life mobile user experience is more related to the AnTuTu result than to geekbench result. But anyway, it's really depends on the user case. If some one prefer to use their mobile phone to simulate the nuclear reaction rather than browser the internet or play some game, geekbench is their first choice to judge their phones:) I really admire the develop who gives geekbench it's name:)And just as a opposite to you, that's why I don't trust the Geekbench result for my mobile choice.
Wilco1 - Thursday, June 20, 2013 - link
Geekbench does the traditional benchmarks, so in that way it is a bit like a small variant of SPEC. However what does AnTuTu measure? When I go to their website I cannot find any information on it. All I see is endless announcements about people cheating the scores. That really instills confidence...Given you seem to have a solid trust in AnTuTu, can you explain in more detail what it actually measures? Have you seen the source code? Is it native code, interpreted or JITted? It seems AnTuTu also does memory, integer and FP scores just like Geekbench, so what makes you claim it is more relevant to mobiles?
virtual void - Wednesday, June 19, 2013 - link
One clarification about using Tegra 3 as referents above, Tegra 3 does score higher on Geekbench than Z2670 and way higher than Z2460.Kurge - Wednesday, June 19, 2013 - link
Yeah.. no. Saltwell is a 2 core, in-order processor.Silvermont is a completely different beast - 4 real, out of order cores.
I _do_ expect Silvermont to come close to A15 and in fact go well beyond.
Wilco1 - Wednesday, June 19, 2013 - link
Why would you expect that? Silvermont is just a 2-way limited OoO core, very similar to Cortex-A9. Cortex-A15 on the other hand is very aggressive with a large reordering window, 3-way and supports 2 memory accesses per cycle (vs just 1 for SilverMont). Given that it should be obvious Silvermont will have a lower IPC than Cortex-A15.You'll be disappointed. Even Intel is claiming just a 2.8 speedup with double the cores and a 20% clock boost. So at the same core count and frequency the speedup from Silvermont is only 16%...
ncp - Wednesday, June 19, 2013 - link
Qualcomm will probably keep the power low as they are cheating on benchmarks.They are detecting what application is launched. For a given list of benchmarks they are tweaking CPU governor to go to full power, tweak data base access and anything yet uncovered.
This way they got very competitive scores on known benchmarks.
This is somewhat smart, however other competitor (IC maker) don't do that so get scores more inline with the way the device is configured in order to get a good pert/power compromise.
Some phone makers are also cheating on top of Qualcomm Cheating, at least one.
Note that probably many phone makers using Qualcomm chipset even do not realize they are cheating on behalf of Qualcomm.
krumme - Wednesday, June 19, 2013 - link
Does that include the "at least one" cheating on top of Qualcomm?Wilco1 - Wednesday, June 19, 2013 - link
In what way is setting the CPU governor cheating? You won't get a better score than the maximum possible score, you'll just get it first time. It seems there is far worse cheating going on with eg. Atoms AnTuTu scores, which are impossibly high (A15 has 3 times faster Stream scores and yet less than half the memory score in AnTuTu?!?).qwerty12345 - Friday, June 21, 2013 - link
looks like wilco1 is a paid pro-qcomm and anti-intc talking machineLaststop311 - Wednesday, June 19, 2013 - link
oh man once we hit 14nm soc arm designs we are gonna have insane performance.FramePointer - Wednesday, June 19, 2013 - link
You're naive if you believe you can accurately estimate the performance of Silvermont based on high level architectural details like the ones you have provided. Given how the current generation of Atom is still using in order execution it should be no match to Qualcomm's Krait or even a basic Cortex A9 (which has at least limited OOO execution) if we were using your logic. Anandtech's own analysis has shown that this is not the case at all."Silvermont is just a 2-way limited OoO core, very similar to Cortex-A9. "
And yet even Clovertrail+ crushes the ARM A9 and it is older than Silvermont.
Wilco1 - Thursday, June 20, 2013 - link
Yes you can make accurate predictions based on the microarchitecture, I've got the experience and patents to make that statement. A9 typically crushes much higher clocked Atoms, eg. http://browser.primatelabs.com/geekbench2/compare/... Both are 4 threads btw, so it is a fair comparison, however a 1.6GHz A9 beats a 2GHz Atom on most single threaded benchmarks as well. Again this is entirely expected based on their respective microarchitectures.Anand's analysis is not believable given he only ever runs JavaScript benchmarks, and these are highly dependent on the browser used. Even on the same hardware there are huge variations. To give an example, the S4 Octa scores 679 on SunSpider according to FoneArena, but Anand claims 920 above. That's a huge difference on the same phone - who is right? The 679 number is correct as that is close to what the Exynos 5250 scores in the Chromebook. So until Anand cleans up his benchmarking I would take his scores with a bag of salt.
shodanshok - Thursday, June 20, 2013 - link
If you think that a single A9 core can beat and destroy a single Atom core, you are wrong.From Anand's iPad2 review:
Alongside multitasking is the performance problem. With the original iPad even deleting several emails at a time was a bit choppy, and web page rendering performance needed tons of work. As always Apple does its best to hide the limitations of the platform but I must point out that even the iPad 2 with a pair of ARM Cortex A9s has lower CPU performance than a netbook with a single core Atom. The fact that you can't really tell most of the time is a testament to Apple's software engineering, but it doesn't change reality.
Some more benchmark:
Tegra 250 @ 1.0 GHz
Core: 2
Compiler: GCC 4.4.5
Coremark/MHz: 5,148
Total Score: 5148,01
Threads: 4
Flags: -O3 -funroll-loop -falign=8 -mcpu=cortex-a9 -mfpu=vfpv3-d16 -mfloat-abi=soft
Atom N280 @ 1.68 GHz
Core: 1
Compiler: GCC 4.4.4
Coremark/MHz: 3,187
Total Score: 5353,79
As you can see, _2_ A9 cores nave total throughput equal to a single Atom core. Atom need high clock of curse, bit it was a specific architectural choice from the start.
The truth is that only A15 decisively defeat Atom.
Unfortunely, Atom GPU was really terrible, being crushed by all others.
shodanshok - Thursday, June 20, 2013 - link
I forgot to specify the benchmark used. It is Coremark: http://www.coremark.org/It is a industry standard benchmark with freely available sources.
Wilco1 - Friday, June 21, 2013 - link
Really? Looking at the published results it shows Exynos 4 does 5560 Coremarks/core at 1.4GHz.The fastest per-core Atom result is 2.3 CM/MHz for 1 thread, and 3.3 with Hyperthreading.
Cortex-A9 does 4.0 for 1 thread - so it is 74% faster single threaded, and 21% faster core for core.
So the A9 destroys Atom on CoreMark as well. I am surprised several of you are trying to argue that in-order cores beat out-of-order cores despite the facts.
shodanshok - Friday, June 21, 2013 - link
No, it is incredible how you pretend to extrapolate _precise_ performance numbers from vague arch details.Return to Coremark site, because you misunderstan the benchmark results. The CM/MHz score represent the score of the entire soc - so it don't rule out core count differences. Let see the CM/core score instead and you will find that Atom is in the same field of A9 scores, sometime much better.
Some examples: Atom z520 vs Tegra2 and Atom n2800 vs exynos4 quad.
Please also note that:
- Coremark does not stress l2/memory in any way. This is the only reason why A9 slow memory interface does not interfere here;
- the compiler has enormous importance in it's score.
The real Atom problem was the terrible GPU and companion chipset.
Regards.
Wilco1 - Friday, June 21, 2013 - link
I listed the per core results, as I said A9 is 74% faster single threaded and 21% faster with Hyperthreading enabled. These are results from the EEMBC website, no complex extrapolation involved.Coremark runs mostly in L1, however it does stress the branch predictor seriously. All benchmarks have a major compiler component. Coremark is horrible like pretty much any EEMBC stuff so I don't think it will become popular.
shodanshok - Saturday, June 22, 2013 - link
I can not agree. From CoreBench site:### Comparison 1:
Tegra2 @ 1.00 GHz (2 A9 cores):
Coremark: 5866.39
Coremark/Core: 2933.20
Atom Z520 @ 1.33 GHz (1 Atom Core):
Coremark: 3192.17
Coremark/Core: 3192.17
Atom advantage: 9%
### Comparison 2:
Exynos4 Quad @ 1.4 GHz (4x A9 cores)
Coremark: 22243.00
Coremark/core: 5560.75
Atom N2800 @ 1.86 GHz (2 Atom cores)
Coremark: 12286.90
Coremark/Core: 6143.45
Atom advantage: 10%
### Note:
Why the two A9 and Atom scores are so much different (see Tegra2 vs Exynos and Atom Z530 vs N2800)? The reason lie in the compiler: recent GCC version have greatly improved their efficienty with in-order uarch. Moreover, please also note that the high A9 score (Exynos) was obtained with their specific arm compiler. I am sure that, if benchmarked using Intel C Compiler, the Atom score would be higher.
### Summary:
the Atom core is more than capable to compete against A9. You can argue than Atom has an higher clock, but in phone/tablet environmento clocks don't mean nothing. What is important is performance/watt.
This bring us to the two real Atom's problem:
1) a very low efficiency chipset and low integration. Moorestown (intel first attempt to mobile with Atom) was doomed from the start because it require 4/5 chips to enable a full-featured phone;
2) a very slow GPU (with very bad performance/watt).
Moreover, it is widely understand that A9 OoO engine is a mild implementation only. A15 is much stronger in this reguard, sometime (not too often, anyway) even apporaching AMD Bobcat single-thread performance.
Regards.
Wilco1 - Saturday, June 22, 2013 - link
No - the performance comparisons that are useful are:1. Max score for a SoC - despite running at a far lower clock, in both comparisons A9-based SoCs win by more than 80% in overall performance.
2. Efficiency of a core at the same frequency (IPC) - Without Hyperthreading A9 is 74% faster, with Hyperthreading A9 wins by more than 20%.
Note that your comparison doesn't work. You can't come to a conclusion about A9 vs Atom performance when you compare with wildly different frequencies. Also it means giving Atom the advantage of having 2 threads vs 1 on A9. So to make the comparison fair you need to compare with an equal number of threads or at the same clock.
Yes GCC has improved a lot in recent years, on ARM it has become a reasonable compiler and competitive with ARM's armcc compiler. I don't know how much better ICC would be on Atom, but I suspect the gap is far smaller as well.
A9 is not hugely OoO indeed, just like Silvermont. A15 is aggressive OoO and beats Jaguar.
shodanshok - Saturday, June 22, 2013 - link
No, I don't agree again.You explicitly talket about CortexA9 and Atom uarch, _not_ their SoC implementation.
You can not use the total SoC score as uarch benchmark - simply because it don't rule out differences in cores number. To measure uarch performances you need to do a core-by-core comparison. Let me do an example: using total SoC score, a 4xA9 SoC is faster then 2xA15 one. However, the latter uarch is considerably more advanced.
A very similar argument can be done for frequency: Atom was _from the start_ designed to hit a relatively high-clock, yet low power target. This was deliberately done to exploit Intel 45/32nm HKMG process, which don't scale power down much for lower frequency target. It is simply a question of design targets: for low power chips, you can get (relatively) high-freq _or_ (relatively) high IPC - not both (actually).
So, you must decide: are you comparing uarch of final SoC implementation? Because, from an uarch point, Atom win. From a performance/watt metric, their bare cores tend to be on par. From a final product specification, A9 is way better because there are many high-integrated, low power, low cost SoCs from a multitude of vendors. On contrast, Atom-based SoCs are offered only by Intel and with a much lower integration factor (and higher cost) - until now,where they latest platform begin to be very competitive against older A9 SoC.
The "little problem" is that ARM is shipping with 2x and 4x A15 cores, and against them Atom is a disvantage.
Regards.
Wilco1 - Saturday, June 22, 2013 - link
While Atom was indeed designed for high frequency, A9 reaches higher frequencies: Atom maxes out at 2GHz on 32nm, while A9 does 1.7GHz on 40nm and 2.3GHz on 28nm. So you can't claim a "microarchitecture" win for Atom when you compare against a low clocked A9.Secondly, since you argue that frequency is an important aspect of the microarchitecture, I would argue that core count matters equally. A9 was designed to be simple and small, so it is typically used as a quad-core. On the other hand Atom is a large and complex core which uses Hyperthreading rather than multiple cores. So if you want to do a fair comparison with Hyperthreading enabled then you have to use 2 A9 cores for every Atom core. That's how they have been designed to be used.
What is the difference between a module, a HT enabled core and a dual core? These are just different ways of improving multithreaded performance with different hardware tradeoffs - but to software they all appear identical.
In conclusion: you cannot just pick whatever comparison you want. Either you compare the whole SoC, including its frequency as well as core count, or you compare microarchitectures normalized on core count and frequency. You can't include one but not the other as frequency, core count and TDP are related.
shodanshok - Sunday, June 23, 2013 - link
So, you started about in-order vs OoO and now you are speaking of die size and perm/mm2?1) While CortexA9 was rated for 2 GHz operation, a single A9 core would dissipate more than 2 Watt at this frequency. Atom is not so much different in this reguard. Moreover, can you point me a phone that use a 2 GHz A9 implementation? I bet no.
2) Atom is also MP form the start: it has the same bus unit and MP capability of Netburst uarch. By which metrics these are inferior to the ARM MP implementation?
3) By die size comparison, A9 is clearly better then Atom. However, its performance are lower.
4) HT is simply a smart sharing of some key structure in order to interleave two thread on the same core. You can not count HT as another core. For example, barrel microprocessors can interleave many threads on a single core: Sun T1 can inteleave 4x threads per core, T2 8x core. Do you count T1 as having 32 cores? If so, you are wrong.
Both I and other users pointed you many reviews and benchmarks where Atom is clearly identified as faster then A9. However, you contine to change metrics.
The only benchmark that paint a different picture is Geekbench, which show A9 in the same league as Sandy Bridge. Do you _really_ think this is true? In SPEC benchmarks, SB is quite close to the big, power hungry but powerfull POWER7. Do you really think that A9 is remotely comparable to this core? Really?
I already stated this: if you compare SoCs, well, A9 wins, because there are many well done SoCs based around it. However, from uarch/performance side, Atom wins.
The funny thing is that is now totally irrelevant: A9 is superseeded by A15, and Atom is very near its EOL. Moreover, Jaguar seems to be a very competent table chip.
Regards.
MrPhilo - Sunday, June 23, 2013 - link
Unfair to compare the A9's to Atom. The Tegra 2 was a old revision of A9 while lacking NEON etc. The newer A9 are more fair to compare. Also a single A9 at 2Ghz wont produce 2 watts at all, the 2.3Ghz Tegra 4i would be worse than the A15 if it did. Remember the nm is 28 not the old 40's.shodanshok - Sunday, June 23, 2013 - link
You are right :) The 2 GHz / 2 Watt rating was at 40nm. It was explicity called at ARM site, so I didn't specified that above.Anyway, lacking NEON instructions means nothing for CoreBech, as it don't test vector code.
Regards.
Wilco1 - Sunday, June 23, 2013 - link
Coremark can be vectorized, which one reason why the scores have improved. Tegra 2 was terrible in most respects, Exynos 4 and Calxeda are currently the best A9 implementations until Tegra 4i is released. Obviously it will be very interesting to see how Silvermont will do against Tegra 4i.Wilco1 - Sunday, June 23, 2013 - link
Now we've agreed that A9 can reach similar frequencies then we can normalize on frequency, yes? When you normalize on frequency then in-order vs OoO matters more, which is what I said all along.1. NVidia has shown their 2.3GHz Tegra 4i reference phone which should be on the market soon. The reason you don't see 2GHz A9 phones today is (a) use of older processes and (b) use of quad cores which leads to too much power use at high frequencies. If they were single or dual core then it would be easier to go for 2GHz.
2. Atom only supports dual core (even Centerton is dual core trying to compete with the Calxeda quad core...), and until the recently released K900 I believe all Atom phones were single core. On the other hand most smart phones have been quad-core since 2012.
3. I still don't see where you think that A9 is slower per core than Atom. At the same frequency it is unbeaten even if you give Atom the advantage of Hyperthreading.
4. If you say that a core with HT is just a single core then surely it should look like a single core to software as well and provide a speedup to single-threaded applications? As soon as you enable HT for a multithreaded application then it would be unfair to compare with single threaded performance on a non-HT core. You can't just pick and choose a comparison that gives one core a major advantage over another.
So far I haven't seen any evidence of native benchmarks where Atom is clearly faster. If you do a fair like for like comparison then A9 wins every time. However if you remove the advantages from one core but keep all the advantages for another then you can of course prove whatever you like.
shodanshok - Sunday, June 23, 2013 - link
1) During Atom lifetime as competent product (read: before A15 release), _no_ phone (or tables, to best of my knowledge) shipped with A9 faster then 1/1.2 GHz. Even Scorpion based phones had higher frequencies. Anyway, from the Coremark comparison written above, you can see that normalizing core/frequency lead to similar per-core performance between A9 and Atom. With the difference that Atom shipped at generally higher frequency, even @ 2.0 GHz: http://www.anandtech.com/show/6307/motorola-hits-2...Comparing a 5-year-old Atom to a to-be-shipped Tegra 4i which as tweaked cores seems fair? I think no.
2) For scaling above 2 cores, currently Atom use FSB communication. However, I can not see why you need 4 Atom cores in a phone. Anyway, you have a point here (on-chip communication are much more pratical) ;)
3) HT is a simple, cheap method to increase total tput. You can not consider an HT-enabled Atom as having 2 cores. Even on coremark sites an HT-enabled Atom is considered a one-core processor. I reiterate: how to consider a barrel processor? A many core approach?
4) You can find many benchmark were Atom excels... some example:
a) http://www.anandtech.com/show/5770/lava-xolo-x900-... (OS android - browser benchmark + linpack + other)
b) http://www.anandtech.com/show/6529/busting-the-x86... (OS WinRT / Win8 - many web brower benchmark)
You can not mark all web browser benchmark as "flawed". Sure they are not 100% uarch dependant, but they are a good method to evaluate the entire software stack used by these processors.
Regards.
Wilco1 - Sunday, June 23, 2013 - link
1. Not true - A9 was running at 1.4GHz in 2011 products, 1.5GHz early 2012 and 1.7GHz late 2012. So A9 was most definitely available at higher frequencies, including in phones (One X). With the shrink to 28nm it gets another boost to 2.3GHz. So in frequency it can match or beat Atom (note by the time it comes out, it'll be compared with Silvermont which achieves similar frequencies).3. Yes a barrel processor looks and behaves like multiple slow cores. The fact that they share hardware in a very specific way is an implementation detail. An equivalent approach would be multiple independent cores. Do you consider a Bulldozer module as a single core as well?
If you prefer to just compare a single core then you should only consider its single threaded performance. As soon as you compare multithreaded performance, then it is completely fair to allow each CPU designer to decide how to implement that, whether it is with HT/SMT, multiple cores or AMD's inbetween approach. You're right of course that HT is a fairly cheap way of adding more threads if you already have a complex core - using multiple cores is only a good idea if you designed your core to be as small as possible.
4. Anand is the worst place to look for CPU benchmarks. Most are JavaScript benchmarks which can vary by a factor of 2 even on the same browser. And his Linpack test is more a Dalvik test than about FP performance. Although there are cases where performance of a browser is important (things like rendering pages), we are discussing micro architecture performance, not how the Android software stack performs.
shodanshok - Monday, June 24, 2013 - link
1) Almost all widespread phone had A9 running at 1.0/1.2 Ghz for so much time. Only recently they increased their frequency at about 1.4 Ghz, and atom is @ 2.0 GHz now.2) "Look as" is not the same as "it is". So a T2 is the equivalent of a 128 core machine? Really? You are confusing SMP with SMT. Look at coremark results: a quad core A9 has performance 4x of a single core, and the score/core reflect this (it is the total score divided by 4). Quad thread Atom instead is a two core uarch, only slightly (30%) faster when tested with 4 thread instead of 2. In fact, coremark/core score is obtained dividing the total score by two, even if the test run with 4 threads. Even coremark developers are wrong?
3) All the posted benchmark,with the exception of Geekbench, show how Atom single core performance are better then A9. Yet you refuse to look at the data, dismissing them as "flawed" or inappropriate. Even user experience, in your eyes, does not matter. can you point us to some reliable bench?
Regards.
darkich - Thursday, July 4, 2013 - link
I think Wilco gave up on discussing with you since you just proved your huge ignorance (the A9 clock speed) and are even trying to defend it.When was the One X + released?
What about the Tansformer prime infinity?
darkich - Thursday, July 4, 2013 - link
I think Wilco gave up on discussing with you since you just proved your huge ignorance (the A9 clock speed) and are even trying to defend it.When was the One X + released?
What about the Tansformer prime infinity?
kukzero - Friday, June 21, 2013 - link
absolutely impossible! The 7900GTX can run Crysis in 720p (30fps),but the Snapdragon 800 can run it? hehe wqnmlgbemperius - Friday, June 21, 2013 - link
Wow. Imagine this paired with Sony and Nexus!? YESParhelion69 - Thursday, June 27, 2013 - link
Why is it Android CPU benchmark broken? Please enlighten me.I thought Antutu was a very good benchmark. And probably Geekbench as well.
darkice1111 - Tuesday, July 9, 2013 - link
Great performance. Now if only we could get some more software optimizations on Android... My iPhone 5 iOS 7 beta 3 results: Sunspider 1.0 - 709.0ms; Kraken - 13783.9ms; Octane v1 - 3056; Browsermark 2.0 - 3056. So 9 month old dual core hardware that's faster than anything on the market today, and faster in some benchmarks than something that's not even on the market yet... Google, wasssssssup??sna1970 - Tuesday, July 9, 2013 - link
Hi,How About comparing this to Nvidia Tegra 4 ?
MaxH - Tuesday, October 1, 2013 - link
Thanks Brian for a great test of this chipset. I am especially interested in the video encoding performance, and your inclusion of the 'MediaInfo' screen capture is really useful to see how it is encoding H.264 video.Your MediaInfo clip clearly shows that this chipset can encode H.264/AVC at 2160p @ 25fps @ 120Mbps, (Baseline @Level 5.1). As a low-budget film-maker, I am speculating about the possibility that the encoder could alternatively be configured to handle 1080p @ 30fps (perhaps 60fps) @ 4:2:2 colour sampling @ 10-bit (perhaps 12-bit) depth. I have not been able to get confirmation of this, but if so - at this price point, this chipset could potentially unlock high quality video capture on regular consumer-level DSLR-type cameras; something that has been limited to commercial broadcast cameras (at high-budget prices) up to now. If anyone is familiar enough with AVC profiles and Levels (and related matters) to be able to speculate about this, I would like to hear your thoughts. Thanks again to Brian.
Netwern - Tuesday, December 31, 2013 - link
Meanwhile Apple engineers...