Arm vs Intel Benchmarks

ChipFrequencyGeek Bench 2 32(/64)Geek Bench 3 Single-CoreGeek Bench 3 Multi-CorePassmark
Apple A51782263503
NVIDIA Tegra 31.314793411089
MediaTek MT65891.516063971258
Intel Atom D5251.813784191039728
Intel Atom Z2760 (Clover Trail)1.813554271092679
MediaTek MT65921.6644402355
Intel Atom Z2580 (Clover Trail +)214735421205
Qualcomm S4 Pro1.526506221943
MediaTek MT81351.7/1.22895
Qualcomm Snapdragon 6001.932636741831
Apple A61.215817141296
Apple A6X1.317767701420
Exynos 5 Octa 54101.635967761997
Intel Atom Z3740 (Bay Trail)1.37972630
Exynos 52501.724338701567
Intel Celeron 8471.125008751330982
NVIDIA Tegra 41.838729062603
Qualcomm Snapdragon 8002.1537339162771
Intel N2830 (Bay Trail Celeron)2.169281605
NVIDIA Tegra 41.944939502900
Intel Atom Z3770 (Bay Trail)1.46
2.39 burst
9723095
Exynos 54201.99802723
Qualcomm Snapdragon 8052.710492878
Apple A7 32-bit app1.310781938
Intel Pentium 2117U1.83390123819191820
Intel Celeron 1007U1.512792220
Intel Celeron 2955U1.413502387
Apple A7 64-bit app1.314142564
Intel Pentium 2127U1.916342914
Intel Core i3-3217U1.83590 / 4617146130952288
Intel Core i5-3317U1.74693 / 6497219942723093
Intel Haswell i5-4250U1.324464600
Intel Haswell i7-4770K3.5~12000 / ~1800044051677410016


I have ordered the table above by Geekbench 3 single-core results. Where information is missing, I make a best guess attempt to place it. It should be noted the Geekbench 2 results above are known to disadvantage Intel chips (maybe 5%). The Passmark benchmark is for x86 chips only.

Geekbench result for MediaTek MT8135 is from MediaTek.
Geekbench 3 results for Apple A7 are from:
The Apple A7 32-bit app result is running on a 64-bit version of iOS.

Geekbench 3 results for Qualcomm Snapdragon 805 are from:

Geekbench 3 results for Clover Trail and Bay Trail are from:
Geekbench results are from: http://browser.primatelabs.com/
Passmark results are from: http://www.passmark.com/

Notes:
Especially when dealing with different architectures, take benchmarks with a grain of salt.

I include the Atom D525 here as it was commonly used in netbooks.

I include the Celeron 847 as it is the chip I consider minimal for running Windows, but decent for running Linux.

I include the Intel Pentium 2117U, Pentium 2127U, Celeron 1007U and the Intel Core i3-3217U as they are popular chips in laptops that actually sell. I consider a laptop in this range to be good enough for 80% or more of users.

The Tegra 3 is used in the Microsoft Surface RT and the ASUS Nexus 7 (2012).

The MediaTek MT6589 is used in many mid-level smartphones around the world.

The MediaTek MT8135 will be used in an upcoming entry-level Amazon Kindle Fire.

The Snapdragon S4 Pro is used in the ASUS Nexus 7 (2013).

The Exynos 5250 is used in the Samsung Chromebook and the Nexus 10 tablet.

The Apple A5, A6, A6X and A7 are in the iPad 2, iPhone 5, iPad 4th gen and iPhone 5S, respectively.

The Intel N2830 is used in the Acer C200 Chromebook that gets 12 hours of battery life.

The Intel 2955U is used in the HP 14 and Acer C720 Chromebooks.

The i5-3317U is used in the Microsoft Surface Pro.

The Haswell i5-4250U is used in the 2013 MacBook Air.

The Intel Haswell i7-4770K is the current flagship high-end desktop chip.

41 comments:

  1. you really shouldnt use geekbench2 because of confirmed bugs in x86 code leading to lower score (were fixed in geekbench3)

    ReplyDelete
  2. I will be adding geekbench 3 scores soon... not enough available at the moment.

    As to "confirmed bugs in x86 code..."... I have never heard of such a problem. Could you elaborate? Searching on the web, I can find no mention of such a problem. In practice, I feel confident that the geekbench 2 numbers give a reasonable rough estimate of where processors stand relative to one another.

    ReplyDelete
    Replies
    1. forums.anandtech.com/showpost.php?p=35249270&postcount=46

      this itself comes from the thread of antutu bugs (only there x86 score was inflated). take note of them as well if u didnt already somewhere (but i see u have hardly any antutu results anyway)

      Delete
    2. I've been going through a few Geekbench 3 results, and it does look like Intel chips have been disadvantaged. I had indeed known about the Antutu problems, and thus there are no Intel results currently in the chart.

      This seems to be good news for Bay Trail... assuming power characteristics are as good as expected, Intel might actually have a decent tablet chip.
      Judging from the parts of Geekbench 2 that are not being questioned, a 1.46 Bay Trail lags slightly the Octa 5410 (1.6ghz). The 1.8 Bay Trail could possibly beat the Tegra 4. Again, this all assume that power consumption is good.

      I'm confused about why there has been no Bay Trail device announcements. Perhaps we'll know more September 11th.

      In the mean time, I'll scare up what Geekbench 3 numbers that are available. I am really glad they are doing single core and multicore scoring this time.

      Thanks for pointing out the problem with Geekbench 2.

      Delete
    3. well maybe... but i feel being good as t4 isnt enough. other leaked numbers didnt impress me much either. and even if bt has lower consumption its because of 22nm only. but then 20nm and 16nm arms are coming. intels only hope is to switch to 14nm asap. but then... ;)

      also browsing gb3 results ive found interesting that t4 multicore score (its around 3000) isnt that far off the macair scores already, with a huge difference in tdp.

      now its time to buy popcorn and enjoy the fight 8)

      - Lgk

      Delete
  3. Hello. I think u need to update your table with this x86 result

    http://browser.primatelabs.com/geekbench3/20867

    There are a lot of x86 scores well over the 4000uni/16000multi bar now.
    Likely many have realized that is better to drop the execution of non-essential background tasks before running the test.

    ReplyDelete
    Replies
    1. My usual way of picking the value in the chart is to use an approximation of a median score. Or if only a few are available, use an average. I toss obviously low ones because as you say, some people do not run benchmarks properly.

      The i7-4770 scores are terribly bizzare... ranging from 3112 to the 4935 you've sent. I'm not willing to use that score set as the number as it doesn't really represent the chip well. What is wrong with this chip? I have difficulty believing most people are running benchmarks with significant background tasks going. I could understand a lot of variance on the multicore, but on the single core?

      I do agree that the 3691 I picked originally is too low now.

      My overall goal is to get numbers than have a reasonable reflection of where the chip performs in real life usage.

      Delete
    2. I think that the difference on i7-4770 scores is related to the different kind of cooling systems. IMO some users have a good heatsink and this allow the cpu to run at 3.7Ghz (four cores active) all time....and obviously at 3.9Ghz (two cores active) once the bench allow this. Nothing wrong in this because these are factory fixed turbo frequencies, so the cpu runs between its own official specifics without any kind of oc.
      Moreover some users have fast memory sticks; more bandwidth helps to have a better score in both integer and fp.

      Too bad many desktops have an horrible cooler :(

      Alberto (the other post is mine)

      Delete
    3. The last answer here. When processor gets to really high speed, benchmarks ends up being limited by memory bandwidth. (motherboard, bus, the individual chips, their configuration (1 large or several small), etc. In general most home users have CPUS that are much faster than all their other components. Putting together a balanced system, where no component (CPU, Memory, network, disk) is superior to the other components can easily cut system cost in half vs an eqiuvalent unbalanced system.

      Delete
  4. tegra4 samples:
    browser.primatelabs.com/geekbench3/search?q=arm+roth
    browser.primatelabs.com/geekbench3/search?q=arm+tostab12BA
    browser.primatelabs.com/geekbench3/search?q=arm+maya
    granted, thats too small, but seems consistent

    - Lgk

    ReplyDelete
    Replies
    1. Thanks for the links... you gave me an idea of how to better search for my missing chips, and now I have everything but Clover Trail and Bay Trail.

      Delete
  5. fresh antutu v4 beta is out. looks like completely new scoring formulae. also note sgs4 and k900 relative cpu scores now closer to the latest gb3 multicore scores:
    http://mobiltelefon.ru/i/other/august13/19/antutu_4_tests.jpg

    - Lgk

    ReplyDelete
  6. http://browser.primatelabs.com/geekbench3/52725

    ReplyDelete
    Replies
    1. Thanks for the link! I've added it to the chart above.

      For those not closely following Bay Trail, you should be aware that the 2.39ghz chips is almost certainly going to need a fan and thus will not be suitable for tablets.

      It would however likely make an excellent low-end laptop/all-in-one chip.

      I'm sure I'll have a lot more to say about this in a day or two as Bay Trail officially rolls out.

      Bay Trail usage in tablets is looking very iffy... see:

      http://www.youtube.com/watch?v=aC5oUdscfq4

      and

      http://www.extremetech.com/computing/165808-intels-bay-trail-benchmarked-makes-first-appearance-in-toshiba-8-inch-windows-8-1-tablet

      It appears that the ones that have the lower power consumption necessary, do not have sufficent CPU power to drive Windows 8. They possibly could be used with Android, but frankly they don't look competitive against ARM chips when their TDP is low enough for tablets.

      Delete
    2. im certain that 2.4ghz is for turboost, this bt might as well be the same one with 1.46ghz base freq. also your multicore t4 score appears too low, its more likely around 2900 (there are now more shield results)

      - Lgk

      Delete
    3. Your theory about the turbo boost may be correct. However, doesn't really matter. It is believed that turbo boost sends the chip to a 7.5 Watt TDP, making it unusable in tablets.

      I've changed the Tegra 4 numbers to match the new results.
      Thanks!

      Delete
    4. How can you say that Baytrail is not competitive with ARM chips once the TDP goes down??? z3770 has a 3W TDP and is pretty in line with Snapdragons SKUs at the same power consumption or clock speed.....do not forget that Snapdragon 800 is a pretty power hungry Soc with a >7W power budget in tablets. Not to mention that Geekbench is not SPEC and it is not able to stress the SOC (core, L1,L2,memory controller, memory), what really matters to have the real performance of a device.
      Geekbench results need of grains of salt, the subsets are very short and they run easily in the L2 or in the L1 in the worst case.
      About your last post.....Qualcomm says that Krait 400 scores 1.3W/core, if you add 1W for soc buses and memory controller power consumption, another 2W for the GPU running at the lower clock speed allowed, you have 7/8W for Snap 800 in a tablet (not in a phone !!!! where the cores are slowed down every few milliseconds).
      So yes a 7.5W Soc can be utilized fanless in a Tablet.
      You know that Intel TDP is related to the cores running at nominal speed and the GPU running without turbo..all together, like in a desktop SKU.

      Delete
    5. I did not say that. What I said was that turbo boost seems not to be an option in tablets as it is believed to make the chip TDP go to 7.5 Watts. At this stage, we do not absolutely know that turbo boost takes 7.5 Watts, hopefully we will have firm facts tomorrow. What the leaked numbers for Bay Trail TDPs imply is that only the slower clocked CPU/GPUs meet the 5 Watt level... the known Bay Trail benchmarks at that level are not competitive with upper end ARM chips. Again... hopefully tomorrow we'll get firm facts.

      Intel has done little to get me to trust them. Their stories about Clover Trail were quite fanciful. What I _fear_ (not know) is that they want to use turbo boost to talk about their chips being competitive in speed, but use TDPs without turbo boost to talk about power being competitive. When it comes down to putting one in a tablet, they just won't be competitive with ARM. Certainly the two reviews of Bay Trail tablets I've seen so far(see the links in my comment above) complain of sluggishness and bad battery life.

      Qualcomm has said they target 5 Watts in tablets.
      See: http://www.fudzilla.com/home/item/31532-qualcomm-aims-at-25-to-3w-tdp-for-phones
      At least according to the article, the TDP of the Snapdragon 800 is < 5 Watts

      I've seen others use 5 Watts TDP as the practical limit for tablets as well. Note, you can use higher TDPs with passive cooling techniques, you just generally don't have a good way to do this in a tablet and make it both thin and comfortable to hold. (Again, one of the Bay Trail tablet reviews complained about it being thick.)

      Delete
  7. I'll have more to say about this later in a blog post, but initial at least, it appears the "leaked" Bay Trail TDP numbers were faked. Indeed, I suspect that several of the Geekbench results may be faked. Again, initially, it looks like Bay Trail is competitive with high-end ARM chips. It may actually have better battery life. I have do have one concern about how its burst mode kicking in and out will affect game play.

    ReplyDelete
    Replies
    1. The promised post can be found here:
      http://computingcompendium.blogspot.com/2013/09/bay-trail-intel-is-competitive-with-arm.html

      Delete
    2. also look http://www.ixbt.com/news/hard/index.shtml?17/16/17

      so its good, but if apples 2x performance promises are even half true, the picture suddenly becomes not so bright ;)

      - Lgk

      Delete
    3. I agree, the game is far from over. Personally, I would prefer that neither side "win". It's much better for consumers when there are two or three serious competitors. (The third would likely specialize in some significant niche.)

      One of the reasons that the PC industry is in the serious shape it's in is that Intel and Microsoft managed to lock in unreasonable margins for a long period of time. There seems to be hope for Intel now and it does not surprise me that they are turning around as they always had AMD around to help keep them honest. I'm doubtful about Microsoft's future though: too much baggage and have not really needed to compete for over a decade. The most symbolic and practical example I can give is the fact that Windows 8 still can run MS-DOS programs.

      Delete
  8. There are some missing pieces that are useful in showing the comparisons in performance between ARM and Intel. Specifically using Ivy Bridge and Haswell platforms in both 32 bit and 64 bit categories. Showing those figures points at a larger gap between Core and Atom/A7/ARM.

    ReplyDelete
    Replies
    1. Generally speaking, I limit the chart to those that are usable in phones and tablets. In addition, I've added a few chips that are commonly available in laptops so that one can get a feel for how these mobile chips relate to those of common experience. I chose the ones you see in the list by examining Amazon's best selling laptops list. And finally, I try to have a chip that represent the current state of the art available to normal users. Currently that is the i7-4770K. That's just to give an idea of how slow the mobile chips still are.

      I do have both Haswell and Ivy bridge chips in the list. There is only one Haswell chip currently as the ones for lower end laptops really are not out yet.

      I will definitely be adding the Haswell chips that don't need a fan. Intel demoed a few of these at IDF, but didn't say what the chips were, so I haven't looked for benchmarks yet. Certainly any Haswell chip that makes it's way into a tablet will be included on this list.

      If you have suggestions for specific chips, I'd be glad to see it. One definitely cannot cover all of Intel's desktop chips, they have an annoying habit of having way to many different chips. (It's never clear why.)

      Oh, if people know chip numbers for the Haswell no-fan chips that Intel was showing off, I'd love to know them.

      I was not at IDF and so my knowledge was limited to what I could glean from a live blog.

      Delete
    2. Thanks for your reply. I was really looking at the U Series i5/i7 4250/4650. But on geekbench3 at least it shows scores for 32 and 64 bit single/multi core. I was kind of referring to the A7 32 and 64 bit, because at least for the i5-3317u it looks like the 32 bit score, the 64 bit adds like 200 pts on top. But yea thanks again for the heads up.

      Delete
  9. more a7: http://browser.primatelabs.com/geekbench3/search?q=iphone6%2C2

    - Lgk

    ReplyDelete
    Replies
    1. or just
      browser.primatelabs.com/geekbench3/search?q=iphone6
      (i dont get whats the difference with "6,1")

      Delete
  10. and whats THIS?? O_o
    browser.primatelabs.com/geekbench3/search?dir=desc&q=arm+n%2Fa
    emulated arm? but then isnt such a score too high for that?

    ReplyDelete
  11. "fixed" exynos5 http://browser.primatelabs.com/geekbench3/search?q=universal5420

    - Lgk

    ReplyDelete
  12. Bay Trail's scores are a bit higher, I've seen postings of ST 968 MT 3093.

    ReplyDelete
  13. There has been a lot of fraud w.r.t. Bay Trail benchmarks. The numbers I have are from Anandtech because I trust them to run the benchmarks properly. Once we get a bunch of shipping BayTrail systems, I'll update the numbers as warranted.

    I would take all of the newest processors with more than a grain of salt as there are just not that many results yet.

    Regardless, I never take the highest result, but rather aim at a median-like score that reflects the probable real world performance. These new chips, and especially BayTrail, are going to be a bit tricky, because literally the temperature of the room the benchmarks are run in can be a factor in the benchmark results.
    I guess that has always been true to a small extent, but BayTrail takes this to a whole new level.

    I gave a better description of how I pick a benchmark number in a comment above.

    ReplyDelete
  14. The Geek Bench 3 Single Core difference at 1GHz is now less than 15% between A7 and i7 down from almost 70% between A5 and i3.

    Clearly Apple is catching up with the computational efficiency of the non-mobile class best products from Intel, if we can trust the test. Will be interesting to see if this will lead to a frequency bump for a non-mobile A chip. I don't think so until Microsoft will release Office for the ARM architecture.

    ReplyDelete
    Replies
    1. Office is available on Win RT tablets, which are ARM. Isn't there a version for the ipad as well?

      Delete
    2. A cut down version of Office is available for RT. It's a shame Microsoft doesn't let third parties put Win32 software on ARM, as a version of LibreOffice would be much better than what they are offering. (I've used LibreOffice on ARM under Linux and it is much faster than Office under RT (and fully featured), even on lesser hardware than the the original Surface.)

      The version for the iPad isn't really worth mentioning. It requires a subscription to Office 365 and regardless is rather weak.

      Delete
  15. any idea about the 11.5 watt tdp Pentium 3560Y (haswell?) -- should do better than or equal to the 2955u celeron, and better single core performance than Atom Z3770

    ReplyDelete
  16. check this http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+3560Y+%40+1.20GHz&id=2078

    ReplyDelete

Now allowing anonymous comments (but they are moderated).