The A6 GPU: PowerVR SGX 543MP3?

Apple made a similar "up to 2x" claim for GPU performance. It didn't share any benchmarks, but there are four options here:

1) PowerVR SGX 543MP2 (same as in A5) at 2x the clock speed
2) PowerVR SGX 543MP4 at the same clock as the MP2 in the A5
3) Marginally higher clocked PowerVR SGX 543MP3
4) Next-gen PowerVR Rogue GPU
It's too early for #4. The first option makes sense but you run into the same issues as on the CPU side with higher voltages used to ramp clocks up (also possible that you drop voltages in the move to the new process technology). 
The second option trades voltage for die area, which based on the A5X Apple is clearly willing to spend where necessary.
The third is sort of the best of both worlds. You don't take a huge die area penalty and at the same time don't run at a significantly higher frequency, and you can get to that same 2x value.

The third option is the most elegant and likely what Apple chose here. Remember that overall die size is dictated by the amount of IO you have around the chip. The A5X had four 32-bit LPDDR2 memory controllers, which gave Apple a huge die area to work with. The move to a smaller manufacturing process cuts down the total die area, which means Apple would either have to add a ton of compute (to fill empty space, no sense in shipping a big chip with a bunch of unused area) or reduce the memory interface to compensate. Pair that knowledge with the fact that Apple doesn't have the same memory bandwidth requirements on the iPhone 5 (0.7MP vs. 3.1MP display) and it makes sense that Apple would go for a narrower memory interface with the A6 compared to the A5X.
How much narrower? Phil Schiller mentioned the A6 was 22% smaller than the A5. We can assume this is compared to the 45nm A5 and not the 32nm A5r2, which would mean that we don't have any more memory channels compared to the A5. In other words, it's quite likely the A6 has a 2x32-bit LPDDR2 memory interface once again.

Final Words

There's not much more to add for now. We'll have a device in a week and I suspect the first reviews will be out a day or two before then. Then the real work begins on finding out exactly what Apple has done inside the A6. If anyone has been dying to put together some good low level iOS benchmarks, now is the time to start.
This is a huge deal for Apple. It puts the company in another league when it comes to vertical integration. The risks are higher (ARM's own designs are tested and proven across tons of different devices/platforms) but the payoff is potentially much greater. As Qualcomm discovered, it's far easier to differentiate (and dominate?) if you're shipping IP that's truly unique from what everyone else has.
Now we get to see just how good Apple's CPU team really is.
The A6's CPU


View All Comments

  • techconc - Monday, September 17, 2012 - link

    I'm not quite sure why people feel compelled to compare future versions of product X to current versions of product Y. If that's your best argument, you've already lost.

    Additionally, Apple's competitors seem to be at least a generation behind regarding GPU performance. The Mali based GPUs haven't exactly been breaking any speed records.

    Finally, this may be a shocking revelation, but the moment an off the shelf part becomes better than what Apple has, Apple is free to use that part as well.
  • user11 - Sunday, September 16, 2012 - link


    One question: A6 processor isn't ARM?

    Dual Core or Quad Core? GHZ?

    Thanks for response :)
  • UpSpin - Sunday, September 16, 2012 - link

    Quote Article:
    "The A6 is the first Apple SoC to use its own ARMv7 based processor design. The CPU core(s) aren't based on a vanilla A9 or A15 design from ARM IP, but instead are something of Apple's own creation."

    A6 uses the ARMv7 instruction set and uses ARM IP to build a custom ARM based SoC, just like Qualcomms Snapdragon and Krait processors.
    Tegra, Omap or Exynos SoCs use a vanilla A9/A15 design from ARM, thus they have less to do on their own and less risk that something isn't working properly. Instead of modifying the core as Qualcomm did and Apple does now, they tune the surrounding (more cores, interaction between them, different GPU, different memory controller, different features, ...)
  • tipoo - Sunday, September 16, 2012 - link

    Yes it is ARM. There are two things you can do, licence out the ARM instruction set, or licence out pre-built cores like the ARM Cortex A9 or A15. Apple appears to have done the former this time and made their own core.

    We don't know the core count yet, but if it was a quad they would certainly have said so, as that is easily marketable, money is on a faster dual core. We don't know clock speed and probably never will for sure, we'll just have estimates based on performance. Apple does not explicitly say the clock speed of the iPhones, they only have for the iPad.
  • Lepton87 - Sunday, September 16, 2012 - link

    By all knowledge, knowing that Apple did not increase the Frequency, and it has twice the performance, what would you have guess the CPU was using?"

    How do you know that clock frequency stays at a pathetic 800MHz?
  • tipoo - Sunday, September 16, 2012 - link

    Probably based on frequency/voltage scaling. They could increase the clock speed, but increase it at the same time as reducing power draw even with a shrink is much harder. Reply
  • RogerShepherd - Sunday, September 16, 2012 - link

    I'm very interested in what's in the Apple A6 but when I read this article it comes down to ""I've confirmed that Apple's A6 SoC is based on Apple's own ARM based CPU core" - while the VFPv4 certainly rules out Cortex A9 I can see no technical reason to rule out Cortex A15 or Cortext A7/A15 big-little. BTW as has been pointed out, A7 does support VFPv4-D32, and so could be an ARMv7s processor. More info in Reply
  • Death666Angel - Sunday, September 16, 2012 - link

    "Based on a lot of digging over the past couple of days, and conversations with the right people, I've confirmed that Apple's A6 SoC is based on Apple's own ARM based CPU core and not the Cortex A15." Yes, that is not evidence, but that is sometimes how things are as long as there is no physical device out.
    Also, that blog post is wrong, Anand does say Cortex A7 also supports VFPv4: "Only the Cortex A5, A7 and A15 support the VFPv4 extensions to the ARMv7-A ISA."
  • RogerShepherd - Sunday, September 16, 2012 - link

    I agree that Anand says

    "Only the Cortex A5, A7 and A15 support the VFPv4 extensions to the ARMv7-A ISA"

    but he then goes on

    "If you want to get really specific, the Cortex A5 and A7 implement a 16 register VFPv4 FPU, while the A15 features a 32 register implementation"

    It's this comment that's not correct. In fact, A7 can have a 16 or 32-register version of VFPv4; if it has Neon, it has the 32-register version.
  • Death666Angel - Sunday, September 16, 2012 - link

    That seems to be true. But I don't see how that influences his conclusion that Cortex A7 is not enough of a step up in performance.
    From where I'm standing, A15 is still a possibility, but highly unlikely given the time frame we are talking about. And I have no reason not to trust Anand's word that it isn't A15.

    Honestly, I don't care that much either way right now. I'm not going to buy an iPhone and I will be eating up any hard facts about the SoC regardless of whether its a new design of Apple's take on an A15. It's interesting in either case. :D

Log in

Don't have an account? Sign up now