A6 for iPad 3?

Discussion in 'iPad' started by w00t951, Jun 22, 2011.

  1. w00t951 macrumors 68000


    Jan 6, 2009
    Pittsburgh, PA
    I was thinking, will the iPad 3 be powered by the A5 or the A6? The A5 still has plenty of power for the current resolution of the iPad, but...

    1) If Apple doesn't increase the screen resolution for the iPad 3, there's nothing to draw people to buy the iPad 3 other than a speed boost; the A6.

    2) If Apple does increase the screen resolution for the iPad 3, the A5 doesn't have enough horsepower to push that many pixels. Therefore, the A6.

    What are your thoughts on this?
  2. tnisatard macrumors 6502

    Apr 1, 2008
    I think Apple will just make a minor update to the processor, but still call it A6 to draw people in.
  3. bufffilm Suspended


    May 3, 2011
    not much to think about. i'm sure apple will put in a cpu which will be capable of driving whatever the resolution the ipad3 will have.

  4. Piggie macrumors G3


    Feb 23, 2010
    Apple know that Tegra3 based tablets from other makes should be out by then.

    I would not think they would want to spend the whole of 2012 being shown in tests as slower than Tegra3 machines.
  5. iBaloney macrumors 6502


    Jan 6, 2010
    no retina display was the deal breaker for me on the ipad 2
  6. Michael CM1 macrumors 603

    Feb 4, 2008
    1) Does anybody even know the difference between an A4 and A5? Is A4 single core and A5 dual core? I honestly don't know. Like others said, is it just some marketing thing? Intel kept calling stuff "Pentium" for years even though it was a fancy name for a 586 processor and eventually they went to 686. I have no clue what all the Core i3, i5, i7 crap means. Don't all processors have cores?

    2) Does something called an A6 actually exist, or is this like Justin Bieber promoting 6G?

    3) As the last person I saw mentioned, I'm more interested in display resolution boost. The iPad 2 screen is very very good. But it's not the iPhone 4 screen. I can't imagine the level of awesome of a display at the PPI level of the iPhone 4 on the iPad. I'm especially thinking about this in ebooks.
  7. psonice macrumors 6502a

    Jul 22, 2005
    A4: 1Ghz single core, Arm cortex A8 architecture, imagination SGX535 graphics core.
    A5: 1Ghz dual core, A9 architecture, imagination SGX 543MP2 (dual core) graphics.

    The CPU went from single core to dual core - best cast, double the speed as you have 2 processors instead of one. However, most apps will only use one core and not gain the benefit. They'll still be more responsive though, because the background system stuff (checking mail and so on) no longer interrupts the app, it can work on the 2nd core.

    It also went from A8 to A9 architecture. What that means is that each core is quite a bit faster (maybe 30%?) than the old A4. So you have 2 CPUs, each one is faster than the A4.

    The graphics chip went from the 535 to the 543, which is much faster, and again it went from single to dual core. Apple claims it's 9x faster, in practice it's not always that fast, but it is WAY faster. Unlike the CPU, all apps use both cores on the graphics chip.

    It probably exists in a lab somewhere. It takes a long, long time to go from 'design complete' to being actually manufactured in large quantities, so it's possibly already finished and being tested in a tablet somewhere at apple's HQ.

    I'd love a retina display too, but I wouldn't want things like games running at 1/4 speed (because it would have 4x more pixels to draw!) You need both the A6 and the retina display together, so it can cope with the extra workload the big screen needs.
  8. tsekh macrumors member

    Apr 13, 2010
    Display resolution won't be the only selling point of the next generation iPad.

    NVidia, and other SoC manufacturers are ready to put out quad-core ARM SoC within this year, or early next year. It is safe to assume Apple would be ready to do the same if iPad 3 were to release a year after iPad 2 launch.

    You may think that no one needs a quad-core on a iPad. That may be true for the current iPad/iOS. iOS 4/5 is designed for A4/A5, it is optimized to use on these hardwares. Unlike other platforms, Apple has complete control over both hardware and software, so they can make sure that the software features that they put into iOS runs fast enough with current hardware.

    But better hardware allows Apple to put more power features onto the iPad, and eventually bridging the functionality between iOS devices and OS X devices. When Apple is ready to put out a quad-core iPad 3, iOS 5.x or iOS 6 will be designed with that in mind, probably with more desktop like multitasking and tabbing to take advantage of the extra horse power.
  9. JulianL macrumors 65816

    Feb 2, 2010
    London, UK
    Re does the A6 exist yet, that really depends on what the A6 is of course. If Apple stick to the A4/A5 pattern (if one can call two data points a pattern!), i.e. approximately yearly spring releases, then I suspect that the A6 doesn't exist yet. Pretty much all of the design cycle for modern CPUs is done via computer simulation and the delay between finalising the design(*) and going into production is fairly short, e.g. this AMD announcement (http://www.softpedia.com/reviews/wi...to-be-Launched-in-Q3-2011-Review-184235.shtml) with tapeout(*) and mass production in the same quarter and that part will almost certainly have a higher transistor count than the A6.

    - Julian

    (*) "Tapeout" is the term used in the industry to describe the point when the design is frozen and sent to production, although first tapeout typically goes to a very small run of prototypes to get back first silicon to test before sending the design to the foundry for mass production. The term "tapeout" is from the old days when the design would literally be written out to a mag tape to deliver to the production facility. Nowadays computer design and simulation tools are so good that it is quite rare for first silicon to not work pretty much perfectly. Even in 1985 when the very first ARM tapeout was done (and yes, that was literally on mag tape), first silicon came back virtually perfect. From memory there was one small timing issue on an external interface that, although it could be corrected by external circuitry, was corrected in the second iteration to avoid increasing the component count on the PCB.

Share This Page