First benchmark indicates A14 is major upgrade from A13
Benchmarks for the iPad Air 4 have seemingly surfaced, indicating the A14 Bionic expected to be used in the "iPhone 12" range offers more performance than the high-powered A13 Bionic and the iPad Pro's A12Z Bionic.
The A14 SoC
Apple is due to launch the "iPhone 12" in the near future, and the fourth-generation iPad Air hasn't gone on sale yet, meaning no-one has been able to try out the latest tablet model to see how responsive the new A14 system-on-chip is in use. In benchmarks spotted by Twitter user "Ice Universe" offers what could be the first benchmark for the tablet, and a first glimpse of what could be powering the 2020 iPhone models.
The single GeekBench benchmark for an "iPad13,2" from October 2 indicates it is a tablet with the motherboard number J308AP. Serial leaker "l0vetodream" pointed out the J308AP refers to the iPad Air 4 with cellular, rather than the J307AP used for the Wi-Fi model.
The benchmark lists the chip as 6-core model with a base frequency of 2.99GHz and 3.66GB of memory. The tablet scored 1,583 points for single-core tests, while for multi-core it achieved 4,198 points, under Geekbench 5.2.3.
Comparing the results against Geekbench's list of iOS and iPadOS devices, the single-core performance is higher than the 1,327 observed in the iPhone 11 Pro, which uses the A13 Bionic at 2.7Ghz. On the multi-core side, the A14 also outpaces the 3,300 the A13 Bionic achieves in the same test, but is still beaten by the A12Z Bionic used in the 12.9-inch iPad Pro, which scored 4,644 points.
According to Apple, the A14 Bionic offers a 30% boost for CPU performance, while using a new four-core graphics architecture for a 30% faster graphics boost, compared against the A12 Bionic used in the iPad Air 3. Against the A13, the benchmarks suggest the A14 offers a 19% improvement in CPU performance and 27% for graphics.
Of course, this all depends on whether the spotted benchmark is genuine. With the imminent arrival of the iPad Air 4, more benchmarks are likely to appear in short order, giving a more accurate assessment of the A14's improvements over the A13.
The A14 SoC
Apple is due to launch the "iPhone 12" in the near future, and the fourth-generation iPad Air hasn't gone on sale yet, meaning no-one has been able to try out the latest tablet model to see how responsive the new A14 system-on-chip is in use. In benchmarks spotted by Twitter user "Ice Universe" offers what could be the first benchmark for the tablet, and a first glimpse of what could be powering the 2020 iPhone models.
The single GeekBench benchmark for an "iPad13,2" from October 2 indicates it is a tablet with the motherboard number J308AP. Serial leaker "l0vetodream" pointed out the J308AP refers to the iPad Air 4 with cellular, rather than the J307AP used for the Wi-Fi model.
The benchmark lists the chip as 6-core model with a base frequency of 2.99GHz and 3.66GB of memory. The tablet scored 1,583 points for single-core tests, while for multi-core it achieved 4,198 points, under Geekbench 5.2.3.
Comparing the results against Geekbench's list of iOS and iPadOS devices, the single-core performance is higher than the 1,327 observed in the iPhone 11 Pro, which uses the A13 Bionic at 2.7Ghz. On the multi-core side, the A14 also outpaces the 3,300 the A13 Bionic achieves in the same test, but is still beaten by the A12Z Bionic used in the 12.9-inch iPad Pro, which scored 4,644 points.
According to Apple, the A14 Bionic offers a 30% boost for CPU performance, while using a new four-core graphics architecture for a 30% faster graphics boost, compared against the A12 Bionic used in the iPad Air 3. Against the A13, the benchmarks suggest the A14 offers a 19% improvement in CPU performance and 27% for graphics.
Of course, this all depends on whether the spotted benchmark is genuine. With the imminent arrival of the iPad Air 4, more benchmarks are likely to appear in short order, giving a more accurate assessment of the A14's improvements over the A13.
Comments
A few things.
It's not that a 2010 vs. 2020 iPad can both run an email app adequately. It's that in 2020, consumers want to be running 10x more apps and have 10x more functionality available to them. Think about Siri analyzing your device usage, communicating with a growing number of wireless accessories (like Apple Watch), converting written text to typed text... this all is new functionality that didn't exist (as prominently) in 2010.
Graphene is one tech that's in the works.
https://www.extremetech.com/extreme/175727-ibm-builds-graphene-chip-thats-10000-times-faster-using-standard-cmos-processes
The multi-core is a lot closer although you have to consider that the SD865 runs on 8 cores vs 6 on the Apple A14 (Apple has been using 6 cores on the regular A series chip since the A10- the “X” version for the iPad Pro uses 8 cores). For the multi-core the SD865’s GB5 score is around 3,280 vs 4,198 for the A12.
... which brings me to systems.
Speeding up the CPU or GPU by, for example 30%, is not going to speed up the system by 30%. The CPU, GPU, memory, etc., are only parts of the system performance equation, and the whole system performance is what really matters. Like any system or complex process there is something called the theory of constraints, or ToC. What ToC tells us is that every system has something about it that is the largest or greatest constraint on the overall system performance (or whatever quality measure you care about). The key word here is largest, because it's not the only constraint.
So what happens when you put on your engineer hat and attack the top constraint on the system's overall performance and make it better? Easy answer, something else becomes the new top constraint. Constraints will still be there holding you back, but generally you can't get rid of all of them and should probably focus on the constraints that have the greatest contribution to system performance, those that can be improved cost effectively, and those that are valued by customers even if the improvements are incremental.
Apple obviously understands what contributes to their systems' overall performance and how ToC applies to every change they make. With Apple Silicon they have something in-hand that they can make better with every release, something that appears to be cost effective, and based on customer loyalty and demand for new Apple products, something that delivers customer value. Why would Apple not play that card? But unless you run CPU benchmarks for a living, don't get overly enamored with the raw CPU numbers and benchmarks. If you're really looking for meaningful performance improvements that matter to you, understanding what system constraints are holding you back, a little self administered ToC, is probably not a bad idea. You may be better off spending your money on a faster storage system, e.g., SSD vs HDD, than paying for a faster CPU, or paying for a faster CPU with integrated graphics rather than a pricey discrete GPU.
My iPhone 7 with its 4 generations back A10 processor seems to be chugging along just fine. To be honest, I don't see what "faster" would get me.
But then software, especially on the iPad, will keep getting more and more demanding. So, when buying new, you might want to allow for some reserve horsepower for down the road. It's a bit like 5G -- you're investing mostly in capacity for the future.
"Some of the games in development will also apparently require an A13 Bionic or later version to run, suggesting they will only be playable by newer iPhone and iPad hardware.