Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
A16 in iPhone 14 Pro is 17% Faster Than A15 in iPhone 13 Pro in New Benchmark
- Thread starter MacRumors
- Start date
- Sort by reaction score
How can a comparison of an old Apple device and a new Apple device be biased in favor of anything?
Given the past few generations‘ increases has mostly been in the 10-20% range, how is it not an outlier? The last time we saw a 40-50% increase was back in a9 to a10.
It's worth noting that Apple's biggest performance gains come from when they both reduce node size and do a core redesign. The jump from the 7nm A12 to the 7nm A13 was about 20%, versus 40% from the A13 to the 5nm A14. From what I've read that TSMC's 4nm process right now isn't much better than the second general 5NP node that the A15/M2 are based off of, hence the modest gains. We should see a sizable jump again at 3nm but we're also reaching the point where smartphone chips (and chips in general) are reaching the limits of what physics can allow so the gains will naturally slow, just as we've seen in desktop chips.
Most people will be upgrading from 2 year old or 3 year old phones, in which case they will see a 70% - 110% increase.
Is it though? The number of processing elements are the same as on A14/A15, so unless there is a major architectural update hising there (highly unlikely) I can’t see why A16 would offer much of an improvement. We can probably expect a more significant step on next years A17 SoC on 3nm
Have never needed to do that, going back to around 2008. The only time a web page has been slow-loading was when WiFi/Cellular was poor.
There was also a time when Intel gave us more then double performance every year (Pentium 1-2-3-4). A16 is unimpressive since it looks like there is no architecture change at all. Process change alone gave them 8% boost in frequency (and probably better cooling solution) and this is exactly what we get, 8% performance increase. ZERO IPC performance increase.
My iPhone 13 Pro Max scores 1744 single core and its known to have terrible cooling (Android phones have all kinds of vapor chambers, while 13 Pro has nothing). So better cooling alone would give couple percent better score.
Last edited:
Kind of of a side note, went from an iPad Air 4th Gen to the 5th Gen. I really thought the M1 would blow me away.
it didn't.
That's why I decided to wait for the iPhone 16 or Beyond. Nothing Apples making has that wow, I need this in my life factor.
it didn't.
That's why I decided to wait for the iPhone 16 or Beyond. Nothing Apples making has that wow, I need this in my life factor.
That’s a good point. If Apple talks up the A16 a lot people might be less interested in the regular iPhone 14.
Assuming these numbers are close, and ignoring minor differences between the regular phones and Pro Max phones, single-core YOY:
Multi-core YOY:
Multi-core YOY:
Is your concern over usb speed because you are a developer and send binaries to your phone to debug for development or is there another reason? I ask because I haven't hooked my iPhone up to a wire in years. I don't do iOS development.
I’m just in for the purple. The pill is neat too. Brighter screen is ever so slightly useful. I guess they also get to better leverage the ML cores with the 48mp camera and pixel binning. It’s all just little improvements. For me the main reason I upgrade every year is I send my old phones to my mom and dad alternating each year so I always have a current gen and they never have a phone more than 2 years old.
As to the benchmarks, part of the improvement is just from the increased base clock from 3.23 -> 3.46, I’m guessing the other part is coming from the increased memory bandwidth and maybe some IPC boosts while they redid the chip in 4nm. But let’s be real, iPhones haven’t had any real speed issues since like A11 or A12 or something. Android will never be able to match iOS performance because apple is able to move waaaaaay more api calls into native IL calls. Apple chips are just as much ASICS as they are general cpus, but android just can’t do the same cause it has to support so many different chips. Same thing with windows or even just x86, legacy support and broad component support along with less OS optimized hardware makes them all inherently slower.
People also gotta remember this is only going from 5nm->4nm and is likely almost the exact same chip otherwise. They did it more to make sure they could limit the power draw when it’s in its 1hz refresh mode for always on display.
At least this one is a bit of a redesign. I can’t quite imagine what the next iPad pros are gonna possibly need a faster cpu cause the M1 in my current one is already clearly overkill. Even my older A14X ipad pro has more than enough power lol. Same with the macbook pro 14/16 upgrades since most of the M2 improvements (like the video encoder/decoders, especially for prores) came from the M1 pro/max to begin with. They aren’t ready to move to OLED screens, but tbh even the current model screens already blow away MOST other laptops out there.
Until iOS/iPadOS actually have a need for a meaningful increase in compute power, all the perf improvements are overkill to begin with. But just looking at the trends in iPadOS and macOS, it’s clear they are slowly iterating into a single product. I for one have long since stopped bothering to care about the paper specs of any of my apple devices.
I do wanna upgrade my 13” M1 MacBook Pro to something, but that’s just because I want the hq prores raw decoder to make it much more viable to edit the 4K/8k footage from my Atomos Ninja.
The most clear example of this is the laughable “improvement“ in specs for the latest watches lol.
That never happened from Intel. The early Pentium 4 was often even slower than high-spec Pentium III's, particularly due to Intel's insistence on RDRAM. And they kept the P6 architecture that the Pentium III used around for a few years as Pentium M before deriving the Core architecture from it, as the Pentium 4's Netburst architecture was an evolutionary dead-end.
I use my phone for emails, browsing, messaging, Netflix and the occasional photo/video edit.
The only reason I have a 13 Pro Max is the camera and screen size. I’m pretty sure I don’t push the A15 even close to what it is able to achieve. The most stress I probably put on the CPU must be when editing a photo or using iMovie.
It’s still amazing what a device this size can achieve and even more amazing to see it improve each year.
The only reason I have a 13 Pro Max is the camera and screen size. I’m pretty sure I don’t push the A15 even close to what it is able to achieve. The most stress I probably put on the CPU must be when editing a photo or using iMovie.
It’s still amazing what a device this size can achieve and even more amazing to see it improve each year.
I’ll be upgrading from my 11 Pro (Geekbench 1328/3414 according to Mactracker), expecting a significant performance jump, much better camera, quadruple of space (256 instead of only 64 GB), a brand new battery, and more.
That said, the 11 Pro is still doing a wonderful job on the most recent ios 16 beta. The A13 is more than sufficient for ios 16.
The 14 Pro still has the same 10 hz screen as the 13 Pro one! It’s even stated on the Apple website’s 14 Pro page. All the tech bloggers keep running with this nonsense 1 hz rumor that has never been confirmed and it’s baffling that people cannot read the official Apple release where this info has been debunked since yesterday.
Yes, straight line speed does not necessarily serve a direct purpose, but it does serve one you may not notice: if you can achieve the same light compute in less time, you can return to idle faster. Do the same thing with less power is a major impact for mobile devices.
Where did you get the information on what SKU the screen is? There is no tear down yet and if you have inside info into how it's the exact same display, man, I'd love to know it. Or is this just an assumption? Which means you're accusing Apple of outright lying that there's a new part when there isn't? Apple's tech specs published say "up to 120Hz" but don't seem to define a lower limit (10Hz or 1Hz), but the spec sheet given to journalists (now off embargo) *does* say 1Hz...
Given the physical difference in the screen real estate and the differences in addressable pixels driving the new pill/Dynamic Island function, I am inclined to believe that indeed this is *not* the same part, but if you have evidence of something we are all missing, I am sure you'd be lauded as a hero for sharing it!