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.
First A14 Benchmarks from iPad Air 4 Appear Online
- Thread starter MacRumors
- Start date
- Sort by reaction score
Sorry, but what you are saying is for controlled benchmarks comparisons, where you control all variables.
You picking up 2 random tests out of geekbench is not a controlled benchmark comparison, because you have no idea how those numbers were obtained.
As such, and without having and actual controlled comparison, the average scores gives a far better idea of performance.
Actually it also gives a good idea of average behaviour for those devices, since they have different constrains for the SoC.
Anyone looking at this to evaluate Arm Macs: Keep in mind that whenever Apple launches a significantly new product, the success is not measured by the first gen success. They are always in it for the long haul, so the really interesting part of Arm Macs is how they compare three years from now. Nor how they compete in first gen.
It’s not a “single core” advantage, it’s a single “graphics core” advantage. With graphics cores, the difference is not how long an operation takes, but how many frames per second can be displayed. So you will see smoother graphics in demanding video games. Numbers, Word, Candy Crush, you won’t notice a difference.
The Metal Score is impressive. It's 137% higher than A12 and 72% higher than A13 according to Geekbench results. It's lot more than 30% higher GPU performance that Apple stated at WWDC.
A12 5307, A13 7308, A14 12571
It can mean that A14X and A14Z can also be much faster?
A12X 10860, A14X 25725
A12Z 11665, A14Z 27632
A12 with 4 GPU cores scores 5307. A12Z with 8 GPU cores scores 11665. 4 extra cores means 120% performance increase. An A14Z Mac SoC with 24 GPU cores could score 87876 in Metal. That's between Radeon Pro W5700XT and Radeon Pro Vega II.
A12 5307, A13 7308, A14 12571
It can mean that A14X and A14Z can also be much faster?
A12X 10860, A14X 25725
A12Z 11665, A14Z 27632
A12 with 4 GPU cores scores 5307. A12Z with 8 GPU cores scores 11665. 4 extra cores means 120% performance increase. An A14Z Mac SoC with 24 GPU cores could score 87876 in Metal. That's between Radeon Pro W5700XT and Radeon Pro Vega II.
You wouldn’t fit two of these chips, but you would create one chip with more cores. Making two chips work together properly is difficult, much more difficult than building one chip with more cores. For example, Apple doesn’t sell _any_ Mac right now with two chips; the top of the range is one chip with 28 cores.
Building ARM chips with more cores is no big problems, the problem is that they would use too much power for an iPhone or iPad. With _lots_ more cores it becomes a bit more difficult to get the full performance potential, but Apple is not at that point yet.
Last edited:
No, the average scores are not as useful in this context. That's why nobody ever posts those website averages when providing reviews of the devices in question.
FWIW, I just ran GB5 on my wife's iPhone XR, and it gets 1114 / 2717, which is 9% higher than the 2497 average A12 iPhone score that was posted earlier. Room temp 21C.
EDIT:
I went to the Geekbench 5 website, and the average posted for the iPhone XR is 2199. LOL. 🙃 (The 2497 was for the XS.) That means her 2717 score is 24% faster than that posted average.
What is the point of these chips on iphones ? Maibe on ipads where you convert video and do photo manipulation. Evrithing on ios is locked, no emulators, no desktop mode like samsund dex, etc.
I mean every scenario when this kind of cpu and gpu power would be useful is locked up by apple. Its like getting a porche and having a speed limiter of 120Km/h installed in factory and you cant remove.
Seriously last time i had an iphone and upgraded from my old 7 to xr, the speed difference and fluidity was minimal. Other than benchmarks almost nobody is going to see the difference
I mean every scenario when this kind of cpu and gpu power would be useful is locked up by apple. Its like getting a porche and having a speed limiter of 120Km/h installed in factory and you cant remove.
Seriously last time i had an iphone and upgraded from my old 7 to xr, the speed difference and fluidity was minimal. Other than benchmarks almost nobody is going to see the difference
No review posts those average scores because they do controlled benchmark comparisons, they don’t pick some random number from the internet. As far as usefulness, it’s in the absence of a controlled benchmark, and it shows thermal constrain problems.
Your own number shows that the iPhone has a lower multicore score... I think it proves the point of the original comment
That's why you compare average scores with average scores.
The iPad Air 3rd Gen average score is higher than iPhone XS average score. What @PPietra said is 100% correct. Why would you cherry pick out a single score if from both devices the average score of 1000's of tests is available?
I actually think a lot of the increase in Metal score most likely comes from work done by the NPU that now has a whopping 16 cores instead of 8. I read up a bit about what the Metal Compute Test runs for tests and there's a lot of tests where the NPU might come into play and takes over work from the GPU (edge detection, face detection, DoF, etc..).
Could be wrong tho, I'm not an expert.
Because the average scores are notoriously unreliable. They may suggest a general trend, but you can't take them at face value when you're trying to assess SoC speed.
FWIW, I just ran GB5 on my wife's iPhone XR again, and this time it got 1112 / 2788.
If we were to believe the 2199 average that is posted for the XR, that would mean her iPhone XR is performing 27% higher than average, suggesting it is some sort of special unicorn XR. (Hint: It isn't.)
My "cheat" here is that I took it out of its thick silicone case, but is that really a cheat? If anything I suspect that a lot more people use iPhones in cases than they use iPads in cases.
I think that Apple’s logic was that everyone who wanted an iPad Pro already got one. Air 4, which is more expensive than Air 3, yet less expensive than iPad Pro, can bring a nice winter flow of cash. Air 4 is also compatible with Pencil 2 and Magic Keyboard, so more money in. It is a very smart move, IMO. Then come March 2021 they can come up with another iPad Pro update, featuring even more advanced A14 chip, the mini-LED screen and what not. It’s a win-win for both Apple and their customers, I guess.
Hmm, you're right.
I checked the scores for XR/XS and there's a bunch of low scores for the iPhone XR/XS for some reason (people benchmarking on battery saver mode?) that drag the average score down a lot. Just ran a few tests on an A12 phone and have similar scores to you.
I take back what I said earlier about the averages.
One reason is that one person’s performance can be another person’s power efficiency. If say, for a certain generation of iPhone we characterise the performance of its SoC as “100%”(*) and the next generation of the SoC developed for the next generation iPhone can get to ”130%” when running within the same power and thermal envelopes as the “100%” SoC that gives the designer of the new iPhone options. They could deliver a phone with a performance of “130%” or they could make the power envelope within which this fancy new SoC is required to run within more restrictive when it’s used in the next iPhone, essentially (and simplifying somewhat) by limiting max clock speeds & throttling earlier vs the power envelope it was allowed to run within when it got to “130%” so that battery drain is reduced. That power saving can then be used for lots of things that the designers might like to do e.g. add a brighter screen, add some extra hardware sensors, reduce battery size to make it thinner (we’ve definitely seen that last one a few times!) or some combination of the above. And of course the power envelope doesn’t need be reduced so drastically that no performance increase is seen between the 2 generations of iPhone, some performance increase can be left on the table to talk about at the launch, just it doesn’t necessarily have to be all of the “30%” increase.
(*) I put all those silly ”100%” etc performance examples in quotes to acknowledge the absurdity of classifying performance with a single number. Hopefully people will tolerate this absurdity when I acknowledge it as I do and hopefully realise I only did it to attempt to give a more understandable answer.
In that case, it seems you're agreeing with the point I've been making all along, which is that Geekbench scores on iOS can't be used (except in the roughest of senses) to predict real-world performance on MacOS.
After all, if they could, then we would indeed expect the SC (single-core) performance of the AS chip in the first Mac laptops to beat the SC performance of the fastest Intel chips—because, according to GB, even the iPad's A14 is already beating the fastest (non-overclocked) Intel chips: 1583 for the A14, vs. 1474 for the 2.8 GHz/4.7 GHz Intel Core i7-1165G7 Tiger Lake (which beats the 1413 of the 3.7 GHz/5.3 GHz Intel Core i9-10900K Comet Lake).
In other words (again, talking about SC here): The iPad A14 already beats fastest Intel chip on GB. The fact that you don't expect the first AS Mac laptop (which, for thermal reasons, should be faster still) will actually be faster than the fastest Intel chip is tantamount to acknowledging GB scores can't be used to predict real-world performance (at least in this context).
Given this, why do you keep quoting the GB scores as predictors of performance? [I do mention GB scores as well, but always (or nearly always) with the caveat that they may not mean much.]
Yes, the fastest next-gen Intel desktop chip (Rocket Lake) will certainly be faster than that U-series Tiger Lake laptop/ultrabook chip. But that won't release until late 2021, a year after the first AS Mac laptop, so it won't even be available for comparison. And even if it were, the % increase for the AS chip, in going from an iPad to a Mac laptop, should be at least as great.
[For those who may not be familiar with the model designations, I just looked this following up. Please correct me if I'm wrong: The i9-10900K has the fastest SC performance of any 10th-generation Intel chip. The 1165G7 Tiger Lake is a low-powered (U-series) laptop/ultrabook chip from Intel's upcoming 10 nm 11th-generation laptop chips. It will be offered with different TDPs and, at that base frequency, the TDP should be ~25W.]
Last edited:
The iPad lineup is almost always unbalanced.
The iPad was higher specced than the Air until this update and cost less.
Already in this thread, look here, shows what the previous iPad Air 3 Metal score also.
as below
Yes. A11 introduces a second generation performance controller to combine all the cores at once. A10 has no such option
but the next time you disapprove of my post gnasher729 take a look as to why you mentioned in your other post that A14 doesn’t have 6 cores 🤦♂️ (You seriously said something that stupid) despite you also saying the cores are divided into 2 high performance cores and 4 high efficiency cores which is basically 6 cores
that and the fact that Apple’s non X chips have been 6 cores since A11....
yeah. Should have looked that up before saying A14 doesn’t have 6 cores
He’s saying that ARM chips aren’t going to beat the _fastest_ Intel processors for some time - that is the 28 core Mac Pro.
Guess what. Downvoted again. Two fast cores plus 4 slow cores is fundamentally different from six cores. I won’t even try to explain it to you.
I had tried to look for a comment but couldn’t find a mention. Sorry!
Last edited:
Nope, he was responding to a post I wrote about single-core performance. You can take look at the post chain and see for yourself, but here was the actual sequence:
Last edited:
doesn’t matter regardless. It still has 6 cores just unified
and secondly anyone (besides you) with a brain will say that A11-14 has 6 cores (from what I see in this thread and everywhere) regardless if they know the high performance/high effiency parts of it
😁😁
14-16 hours of batter life? I hope so but reality is probably not there.
The ARM chips going in Mac's will be bigger, hotter and take more power. They will have to be to compete with Intel chips. In fact because of emulation they will need to be more powerful. They will not be constrained by the form factors of the iPhone and iPad and that is good. Battery life will also be impacted a 13 or 14inch screen as well.