iPhone 4 processor speed can now be measured: 800MHz

Discussion in 'iPhone' started by Neblinio, Jan 15, 2011.

  1. macrumors newbie

    Neblinio

    Joined:
    Sep 27, 2007
    #1
    As of iOS 4.2.1, "sysctl -a" command is no longer returning value 0 for "hw.cpufrequency". Now you get hw.cpufrequency: 800000000 for iPhone 4 and iPod touch 4G.

    You can try it any time using mobile terminal, SSH or even Sysinfoplus.



    Here's a screenshot I made of Sysinfoplus 1.2 showing my iPhone 4 processor speed.



    [​IMG]
     
  2. macrumors regular

    Joined:
    May 19, 2010
    #2
    its the part that says CPU frequency...

    but is that the actual frequency, or the underlocked frequency... apple has been known to underlock processor before.

    iPhone and iPhone 3G
    620MHz underlocket to 412Mhz

    iPhone 3GS
    833MHz underlocket to 600Mhz
     
  3. macrumors demi-god

    kdarling

    Joined:
    Jun 9, 2007
    Location:
    Device engineer 30+ yrs, touchscreens 24+.
    #3
    It could be underclocked from a gazillion MHz, and the actual frequency is still all that matters.

    (The only technical interest in underclocking is that it conserves power and heat.)
     
  4. thread starter macrumors newbie

    Neblinio

    Joined:
    Sep 27, 2007
    #4
    iPhone 4 and iPod touch 4G:
    1GHz underclocked to 800MHz


    We all knew Apple's A4 processor was able to crank up to 1GHz (iPad does). What we didn't know was the actual underclocked speed on the iPhone 4 and iPod 4G (because they are obviously underclocked, if they weren't so, Apple would have advertised their 1GHz speed)
     
  5. macrumors 65816

    nunes013

    Joined:
    May 24, 2010
    Location:
    Connecticut
    #5
    for 800 mhz it runs pretty smoothly compared to a lot of these other devices that run at 1 ghz. i wonder what kind of battery we would see if we cranked it up to 1 ghz
     
  6. macrumors 6502

    Joined:
    Mar 18, 2010
    #6
    Looks like "1 GHz processor" may be part of the selling point for the summer upgrade.
     
  7. macrumors 6502a

    Ashin

    Joined:
    Jun 19, 2010
    #7
    Would have been easier to just check Wikipedia
     
  8. macrumors 6502

    Joined:
    Jan 6, 2010
    #8
    Surprisingly its faster than any Android. It crushes my Evo.
     
  9. thread starter macrumors newbie

    Neblinio

    Joined:
    Sep 27, 2007
    #9
    Wikipedia has never been able to confirm its actual clock speed:
    "The clock speed in the iPhone 4 has not been disclosed."

    I remember reading somewhere that its clock speed was somewhere near 800-850MHz, but this time it's confirmed by "sysctl" command.
     
  10. macrumors 6502

    Joined:
    Jan 14, 2011
    #10
    Even at 2GHz the upcoming Bionic and Atrix will lag. What most people do not realize is that the software is more important than the hardware. Android has multitasking but it does not handle it fluidly. As an owner of a Motorola Droid, Droid X, Samsung Fascinate, and HTC Droid Incredible I can tell you that they all force close often.

    Take the browser for example. Most Android phones render the page in totality. If you render Engadget it will bring up every article at the same time; this results in an extremely laggy and jittery browsing experience.

    The iPhone on the other hand renders whatever is on the screen (that's why if you scroll down or up quickly you will see the checkered loading). Consequently the browser is much "snappier" and pinch and zoom is flawless.

    Add to that the fact that the iPhone does not do complete multitasking and lacks the ability to add widgets, therefore it is such a quick phone.

    For the Android handsets to run as smooth as the iPhone the hardware needs to be much more powerful. I feel they need at least 2GHz of processing power, and a 2GB ram. Until then you will see Android lag as before. Unfortunately, with the above specs you will need a monster battery. Lithium will not do anymore so we will need to wait for battery manufacturers to create something smaller and more efficient.
     
  11. macrumors newbie

    Joined:
    Apr 8, 2011
    #11
    How did you get Sysinfoplus to measure your clock speed? I have an iPhone 4 on iOS 4.2.1, I recently installed Sysinfoplus, and it read it as 0.
     
  12. macrumors 68020

    dccorona

    Joined:
    Jun 12, 2008
    #12
    it is underclocked. The A4 processor is identical to the one used in the iPad, but on the iPad it runs at 1ghz
     
  13. macrumors 68020

    dccorona

    Joined:
    Jun 12, 2008
    #13
    thats because of the software...iOS is much better optimized than android devices...Android is great on the outside but the actual OS is just a clunky emulator built on top of linux, and it is almost never specifically optimized for the phone it runs on

    this is not true, however, of WP7 devices, all of which are 1ghz, and all of which (well, the 2 ive owned at least, HTC surround and samsung focus) run very smoothly, as smooth if not smoother than the iPhone 4
     
  14. macrumors 68020

    dccorona

    Joined:
    Jun 12, 2008
    #14
    the android phones suffer, true, but not so much from the reasons you stated, but because of the nature of the OS. The entire thing...apps, widgets, springboard, everything, is running inside of a Java emulator shell over a linux based root OS. None of it is optimized for the specific device (except the nexus 1 and S devices), and (at least from what i read) it is not coded nearly as well as iOS is
     
  15. infidel69, Apr 9, 2011
    Last edited: Apr 9, 2011

    macrumors regular

    Joined:
    Feb 25, 2011
    #15
    It better be a dual core 1+ GHz processor or people are going to be pissed.

    I haven't noticed any lag with my Atrix. Compared to the iphone4 the Atrix is a speed demon anyone who says otherwise is in denial. Look at the youtube comparisons if you don't believe me on this. This is with Froyo which doesn't make much use of multiple cores. When the Atrix gets GB in a few months we should see a nice boost in performance for sure. dccorona and fugimself are just talkin out their asses.
     
  16. macrumors 6502a

    Joined:
    Feb 23, 2008
    #16
    iPhone 4, whatever speed the CPU runs, is fast. Very fast. It still holds it own against the newer 1GHz+ chips and we know Apple will put the dual-core 1GHz A5 chip from iPad 2 in the new iPhone 5.

    It really all comes down to software, as has been said.
     
  17. macrumors demi-god

    kdarling

    Joined:
    Jun 9, 2007
    Location:
    Device engineer 30+ yrs, touchscreens 24+.
    #17
    Or they simply take the chips that tested okay at 1GHz and use them in the iPad, and take the ones that only tested up to 800MHz and use them in the iPhone.

    Use the same chip in a range of devices, from iPod to iPad, and you can save all sorts of money by throwing away less production output.
     
  18. macrumors 6502

    iAmYou

    Joined:
    Jun 29, 2010
  19. macrumors 604

    SnowLeopard2008

    Joined:
    Jul 4, 2008
    Location:
    Silicon Valley
    #19
    It's not really the frequency that matters. A 3.0GHz Pentium 4 isn't going to crush a 2.0GHz Core 2 Duo. The frequency does matter, but it's not the only info you should consider when looking at a device's performance. There are other factors. Just something to keep in mind. Apple does have a history of customizing things to deliver better battery performance while not killing the overall CPU/GPU performance. Tradeoffs you know?
     
  20. macrumors 604

    SnowLeopard2008

    Joined:
    Jul 4, 2008
    Location:
    Silicon Valley
    #20
    Very good point. This is one of the strategies that Intel uses. I remember some of those Celeron and other low end chips were basically high end chips with small defects in them. All they had to do was permanently disable or physically cut off the defected sections and rebrand them as low end chips.
     
  21. macrumors demi-god

    kdarling

    Joined:
    Jun 9, 2007
    Location:
    Device engineer 30+ yrs, touchscreens 24+.
    #21
    That's true, they do that with memory chips as well.

    For example, for the inexpensive type of Flash that Apple uses, each 8GB can have up to 200MB bad from the factory and still be considered "okay". Less than that, and they could set/sell it as a smaller memory if they wished.

    Which I find interesting, because one person's 32GB phone could really have 32GB, while the next person could have 32GB - 800MB = 31.2GB. Sometimes I think that's why Apple goes ahead and chops off the extra space just to hide the differences. (It's definitely not about that decimal vs binary counting nonsense, as Flash chips only come in binary powers.)
     
  22. macrumors 6502a

    Joined:
    Feb 2, 2010
    Location:
    London, UK
    #22
    The most obvious example might be AMD's triple-core Phenom parts which are cast-offs from the quad cores that have a failure localised to a single core which is then disabled to give an X3 that passes the quality tests. In the wild it is even possible to re-enable the disabled core and in some cases, where the failure was marginal on the faulty core, people have run it quad core (http://guru3d.com/news/phenom-ii-x3--enable-the-4th-core/).

    - Julian
     
  23. macrumors 6502a

    Joined:
    Feb 2, 2010
    Location:
    London, UK
    #23
    Didn't some fairly recent generation of one of Nvidea's GPU architectures have 240 shaders/cores? To me as a computer scientist that is a strange and ugly number. I saw speculation that the design had 256 shaders but that the spec was set at 240 so that they could tollerate up to 16 dead shaders on a part before they had to reject it completely. I suspect that manufacturers do detailed modelling in these cases to find the right balance of how far to deliberately cripple good parts in order to increase overall yield, particularly important in high-volume cost sensitive parts since raising the yield reduces the cost per part.

    - Julian
     
  24. macrumors newbie

    Joined:
    Aug 21, 2008
    #24
    CPU dynamic frequency

    Is there a way to read the actual CPU frequency at any given moment? It's my understanding that the frequency is dynamically adjusted by iOS.
     
  25. macrumors P6

    Intell

    Joined:
    Jan 24, 2010
    Location:
    Inside
    #25
    No
     

Share This Page