The iPhone 5's A6 SoC: Not A15 or A9, a Custom Apple Core Instead
by Anand Lal Shimpi on September 15, 2012 2:15 PM EST- Posted in
- Smartphones
- Apple
- Mobile
- SoCs
- iPhone 5
When Apple announced the iPhone 5, Phil Schiller officially announced what had leaked several days earlier: the phone is powered by Apple's new A6 SoC.
As always, Apple didn't announce clock speeds, CPU microarchitecture, memory bandwidth or GPU details. It did however give us an indication of expected CPU performance:
Prior to the announcement we speculated the iPhone 5's SoC would simply be a higher clocked version of the 32nm A5r2 used in the iPad 2,4. After all, Apple seems to like saving major architecture shifts for the iPad.
However, just prior to the announcement I received some information pointing to a move away from the ARM Cortex A9 used in the A5. Given Apple's reliance on fully licensed ARM cores in the past, the expected performance gains and unpublishable information that started all of this I concluded Apple's A6 SoC likely featured two ARM Cortex A15 cores.
It turns out I was wrong. But pleasantly surprised.
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.
Hints in Xcode 4.5
The iPhone 5 will ship with and only run iOS 6.0. To coincide with the launch of iOS 6.0, Apple has seeded developers with a newer version of its development tools. Xcode 4.5 makes two major changes: it drops support for the ARMv6 ISA (used by the ARM11 core in the iPhone 2G and iPhone 3G), keeps support for ARMv7 (used by modern ARM cores) and it adds support for a new architecture target designed to support the new A6 SoC: armv7s.
What's the main difference between the armv7 and armv7s architecture targets for the LLVM C compiler? The presence of VFPv4 support. The armv7s target supports it, the v7 target doesn't. Why does this matter?
Only the Cortex A5, A7 and A15 support the VFPv4 extensions to the ARMv7-A ISA. The Cortex A8 and A9 top out at VFPv3. 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. The point is, if your architecture supports VFPv4 then it isn't a Cortex A8 or A9.
It's pretty easy to dismiss the A5 and A7 as neither of those architectures is significantly faster than the Cortex A9 used in Apple's A5. The obvious conclusion then is Apple implemented a pair of A15s in its A6 SoC.
For unpublishable reasons, I knew the A6 SoC wasn't based on ARM's Cortex A9, but I immediately assumed that the only other option was the Cortex A15. I foolishly cast aside the other major possibility: an Apple developed ARMv7 processor core.
Balancing Battery Life and Performance
There are two types of ARM licensees: those who license a specific processor core (e.g. Cortex A8, A9, A15), and those who license an ARM instruction set architecture for custom implementation (e.g. ARMv7 ISA). For a long time it's been known that Apple has both types of licenses. Qualcomm is in a similar situation; it licenses individual ARM cores for use in some SoCs (e.g. the MSM8x25/Snapdragon S4 Play uses ARM Cortex A5s) as well as licenses the ARM instruction set for use by its own processors (e.g. Scorpion/Krait implement in the ARMv7 ISA).
For a while now I'd heard that Apple was working on its own ARM based CPU core, but last I heard Apple was having issues making it work. I assumed that it was too early for Apple's own design to be ready. It turns out that it's not. 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.
Implementing VFPv4 tells us that this isn't simply another Cortex A9 design targeted at higher clocks. If I had to guess, I would assume Apple did something similar to Qualcomm this generation: go wider without going substantially deeper. Remember Qualcomm moved from a dual-issue mostly in-order architecture to a three-wide out-of-order machine with Krait. ARM went from two-wide OoO to three-wide OoO but in the process also heavily pursued clock speed by dramatically increasing the depth of the machine.
The deeper machine plus much wider front end and execution engines drives both power and performance up. Rumor has it that the original design goal for ARM's Cortex A15 was servers, and it's only through big.LITTLE (or other clever techniques) that the A15 would be suitable for smartphones. Given Apple's intense focus on power consumption, skipping the A15 would make sense but performance still had to improve.
Why not just run the Cortex A9 cores from Apple's A5 at higher frequencies? It's tempting, after all that's what many others have done in the space, but sub-optimal from a design perspective. As we learned during the Pentium 4 days, simply relying on frequency scaling to deliver generational performance improvements results in reduced power efficiency over the long run.
To push frequency you have to push voltage, which has an exponential impact on power consumption. Running your cores as close as possible to their minimum voltage is ideal for battery life. The right approach to scaling CPU performance is a combination of increasing architectural efficiency (instructions executed per clock goes up), multithreading and conservative frequency scaling. Remember that in 2005 Intel hit 3.73GHz with the Pentium Extreme Edition. Seven years later Intel's fastest client CPU only runs at 3.5GHz (3.9GHz with turbo) but has four times the cores and up to 3x the single threaded performance. Architecture, not just frequency, must improve over time.
At its keynote, Apple promised longer battery life and 2x better CPU performance. It's clear that the A6 moved to 32nm but it's impossible to extract 2x better performance from the same CPU architecture while improving battery life over only a single process node shrink.
Despite all of this, had it not been for some external confirmation, I would've probably settled on a pair of higher clocked A9s as the likely option for the A6. In fact, higher clocked A9s was what we originally claimed would be in the iPhone 5 in our NFC post.
I should probably give Apple's CPU team more credit in the future.
The bad news is I have no details on the design of Apple's custom core. Despite Apple's willingness to spend on die area, I believe an A15/Krait class CPU core is a likely target. Slightly wider front end, more execution resources, more flexible OoO execution engine, deeper buffers, bigger windows, etc... Support for VFPv4 guarantees a bigger core size than the Cortex A9, it only makes sense that Apple would push the envelope everywhere else as well. I'm particularly interested in frequency targets and whether there's any clever dynamic clock work happening. Someone needs to run Geekbench on an iPhone 5 pronto.
I also have no indication how many cores there are. I am assuming two but Apple was careful not to report core count (as it has in the past). We'll get more details as we get our hands on devices in a week. I'm really interested to see what happens once Chipworks and UBM go to town on the A6.
Facebook
Twitter
RSS
163 Comments
View All Comments
tipoo - Sunday, September 16, 2012 - link
Look one post above you, it seems to be in the ballpark of A15, but A15 isn't out yet so I guess we'll see.tru_th - Monday, September 17, 2012 - link
For those of you referencing the Geekbench 2 Results where the iPhone5 beats the Galaxy S3, you have been given the incorrect link. The link that shows the Galaxy S3 test with a lower score than an iPhone5 is because you were linked to the wrong test. The test you were linked to is under android tests which is the incorrect score. If you go to the same website and go to the popular results:http://browser.primatelabs.com/geekbench2/popular
You'll notice that the iPhone5 Scores1601, and the Galaxy S3 scores 2059. I also noticed that the Galaxy S3 they were testing only registered ~800mb of RAM when in actuality the S3 has 2GB. . The Galaxy S3's hardware is superior to the iPhone5, and Apple isn't going to release another phone for who knows, a year? They dropped the ball with this one in my opinion.
What hardware is better in the S3 than the iPhone5? Better front facing camera (1.9MP), NFC chip, Larger & Removable battery, additional memory via SD card (up to 64GB), larger & full 720p resolution screen, and there's more, just take a look at Samsung's ad:
http://www.macobserver.com/tmo/article/samsung-gal...
darkcrayon - Monday, September 17, 2012 - link
Direct comparison of geekbench scores between the iPhone 5 and SGS3 isn't very meaningful. However assuming the direct comparison with the iPhone 4S and 3rd gen iPad is more or less valid, then the iPhone 5 is going to have a *ton* of performance headroom and plenty to last until the 5S or 6. And this doesn't even say anything about graphics performance which is going to be the top for a smartphone by quite a large margin if it's 2 times the 4S.The iPhone 4S keeps up perceptually with a Galaxy S III. So we're expecting the 5 to surpass it from a general use perspective.
Also, size is a tradeoff. Saying "because it's bigger!" is not an advantage- or the iPad totally destroys the galaxy S3, right? The SGS3 is better because it's bigger. And the iPhone is better because it's smaller.
tru_th - Monday, September 17, 2012 - link
I believe the comparison between the S3 and the iPhone 5 is valid because they are both at the top of their markets, don't get into tablets that is a completely different subject. The iPhone 5 didn't improve the screen from the 4S, they added height to the top and the screen is not even 720p capable. The S3 is, and many other top of the line androids are. The iPhone 5 finally includes real 4G, while Android has had it for over a year. My point is that Apple is not innovating. You can say its better because its smaller, but because Android is open source, there are many companies like Samsung, HTC, LG, etc all making smartphones there will be smaller Android phones that still are superior to the iPhone 5. Like I said, Apple releases their phones after usually about a year so by the time the 5S comes out Android will be even further ahead, I imagine quad core phones will be the regular by then since the S3 is already a quad core phone. Android might even be optimized for 4 cores by then as well. Apple needs to stop innovation with their ridiculously broad patents and release some quality features for their phones. Keeping up with the hardware standard should be the first thing they improve on. The only thing I was actually impressed with on the iPhone 5 was the 5 microphones for better sound quality, although I'm not even sure it improves it I think it is a good idea. The screen, the processor, the 1 gig ram, the "different" plug that improves nothing at all, and the camera are all areas where there was no significant improvement compared to similar priced phones. Apple has in the past been at least up to standard with their releases, which holds the iPhone users at bay until the next release, but like I said I don't think they achieved that with this phone. I find that even Apple fans are a little let down that Apple only slightly changed the 4S to the 5.tru_th - Monday, September 17, 2012 - link
Opps, I meant "apple needs to stop stifling innovation with their ridiculously broad patents and stick to releasing some quality features for their phones"darkcrayon - Monday, September 17, 2012 - link
I mentioned tablets as a point of comparison, as conceptually they are similar. The reason I said the iPhone 5 and Galaxy S3 comparison wasn't particularly valid was in terms of the Geekbench score- it might not be fair to one side or another because the operating system is different. In the real world, comparing the devices next to each other for *actual* performance is valid. And considering a 4S keeps up or exceeds a SGS3 on doing mundane things like tapping to zoom in a web browser, the iPhone 5 is really going to shine- again- in comparison. And of course iOS still has the most CPU/GPU demanding apps to even take advantage of such a processor (polished, feature rich first party apps - pro audio apps, and then high end games that are not yet available for other mobiles)Apple _is_ innovating- the evidence is this very article. Trying to dodge that with complaints about "patents" is completely beside the point. The only thing that would put Android "ahead" here really is that you can fit a larger, hotter chip inside of a larger device. Hence again my comparison to the iPad.
tru_th - Tuesday, September 18, 2012 - link
I understand the geekbench score not being valid. But until a reputable source tests them both side by side, both running the latest firmware (JellyBean vs iOS), the tests are all we have. The operating systems are indeed way different and its hard to compare them. In terms of hardware, you cannot deny that the S3 is superior. A very large amount of the iPhone 5's parts (processor, memory, screen) are manufactured by Samsung (up to 25% of the phone). As for iOS having the most CPU/GPU demanding apps. where are you getting this information? There really isn't much difference between apps. A large majority of apps are available on both the Google Play store and the App store.Apple is no longer innovating. The new iPhone5 has nothing new to bring to the table aside from having 5 microphones on the phone. All the other additions were just old technology that Apple had yet to add to the phone (4G, larger screen), or expected improvements such as the processor.
Anyways. My main point is that Apple did release a better phone but they did not improve enough. The S3 and iPhone5 are both top of the line phones and it can be argued by either side which one is faster. But the S3 came out before the iPhone5 did and the Android phone makers (Samsung, HTC, Motorolla, LG) all release phones and technology improvements faster than Apple. We might see a couple of new top of the line smartphones from each of the top companies before we will see the iPhone5, and I think by that point Apple is going to be further behind. As for the comment below me that Samsung is "scared", I hope you realize that they sell two phones for every Apple phone sold in addition to making up to 25% of the iPhone and iPad's. I think they're okay.
techconc - Monday, September 17, 2012 - link
There are a couple things to note.First, we have more data points for the GS3 then we do for the iPhone 5.
Second, the Geekbench tests are multi-threaded. Essentially, showing a best case scenario for quad core machines. In real life, it's debatable how often you'd ever see that advantage. Remember, the iPhone 5 is getting this performance with half the number of cores (and running at a 40% slower clock speed), which means each core is significantly faster on Apple's chip.
Third, the iPhone 4s "feels" faster than a GS3. That might be due to poor Android GUI optimizations or reflect upon the 4s' superior GPU performance. Either way, the iPhone 5 doubles the 4s's performance (in GPU also). Which by extension means that the iPhone 5 will feel faster than the GS3 for most tasks.
As for Samsung's ad, these are the actions of someone who is scared.
techconc - Monday, September 17, 2012 - link
One thing I find troubling is the type of benchmarks Apple chose to use. They seem to be more disk i/o bound than CPU specific. Loading applications, saving files, etc...Lucian Armasu - Monday, September 17, 2012 - link
Why is it so hard for people to notice this? Geekbench also tests memory performance, and Anand has already said it has vastly improved memory (which btw, could be a newer memory bought from Samsung).So why the hell is all the focus on the CPU? Even if iPhone 5 has 1600 score and GS3 has 1560, that does NOT mean that it has a faster processor.
Again - the test measures BOTH memory and CPU performance, and it obviously means iPhone 5 has faster memory than GS3, which would give it a higher score for that...which is how the iPhone 5 compensates for actually having a SLOWER CPU.