Other A14 benchmarks

[EugW]

I will leave it up to you to decide if this is legit.

Leaked A14 Bionic AnTuTu score surfaces online; Apple's newest silicon fails to surpass the Qualcomm Snapdragon 865

Apple's latest A14 Bionic chip has just made an appearance on AnTuTu. It nets a total score of 572,333, which is significantly lesser than that of the Qualcomm Snapdragon 865 and Snapdragon 865+. The low number could likely be due to unstable software, so we should take the results with a grain...

www.notebookcheck.net

https://twitter.com/i/web/status/1306543787762544640

 

[JPack]

The scores sound about right as Apple doesn't score well in AnTuTu benchmarks. The situation is similar with 3DMark - high peak scores, but relatively low sustained performance.

 

[EugW]

The scores sound about right as Apple doesn't score well in AnTuTu benchmarks. The situation is similar with 3DMark - high peak scores, but relatively low sustained performance.

OK, so maybe legit. I think the concern by some people wasn't only that the AnTuTu scores were not #1, but more that the improvement over A13 was less than some expected.

I'm still waiting for the Geekbench 5 scores. A13 scores approx. 1340 single-core and 3500 multi-core. I wonder if A14 can hit 1600 and 4100 respectively. In comparison, A12Z gets 1130 and 4700 respectively. A14X is gonna be a monster! Faster single-core than every single Mac, and MacBook Pro Core i9 territory for multi-core.

 

[JPack]

OK, so maybe legit. I think the concern by some people wasn't only that the AnTuTu scores were not #1, but more that the improvement over A13 was less than some expected.

According to AnandTech, they estimate 16% and 8.3% for the CPU and GPU, respectively. Based on that, the CPU numbers match exactly. But the GPU scores fall a bit short. Maybe that's due to AnTuTu being more difficult on Apple silicon.

Apple Announces 5nm A14 SoC - Meagre Upgrades, Or Just Less Power Hungry?

www.anandtech.com

 

[EugW]

According to AnandTech, they estimate 16% and 8.3% for the CPU and GPU, respectively. Based on that, the CPU numbers match exactly. But the GPU scores fall a bit short. Maybe that's due to AnTuTu being more difficult on Apple silicon.

Apple Announces 5nm A14 SoC - Meagre Upgrades, Or Just Less Power Hungry?

www.anandtech.com

I guess that agrees with this page:

TSMC Details 3nm Process Technology: Full Node Scaling for 2H22 Volume Production

www.anandtech.com

If Apple has kept the same power levels, then a 16.7% improvement makes sense. That's right on par with TSMC's +15% estimate.

 

[PeterG727]

I guess it could be due to throttling? Apple never puts any sort of vapour chamber cooling/heat pipes in iphones.

 

[bobmans]

Since UniverseIce mentioned on Twitter that this score is lower than the Snapdragon 865, I'd like to mention that you can not under any circumstances compare Antutu scores from the Android app with the Antutu scores from the iOS app.

I always have a good laugh if people say that Snapdragon chips are faster than the Apple A series chips (or the other way around) and then post an Antutu benchmark.

You can read this Antutu blogpost that explain why they're not comparable here: http://www.antutu.com/en/doc/119646.htm

 

[EugW]

CPU Monkey claims to have Geekbench 5 scores for A14: 1658 / 4068

Apple A14 Bionic - Benchmark, Test and Specs

Apple A14 Bionic - Benchmark, Geekbench 5, Cinebench R20, Cinebench R23, Cinebench R15 and FP32 iGPU (GFLOPS) benchmark results

www.cpu-monkey.com

For comparison, A13 gets about 1340 / 3500.

 

[The Game 161]

With SD865+ and A14 both are great chip sets you won’t notice much difference in day to day use

Sometimes people pay too much attention to benchmarks.

 

[Coffee_Time]

Since UniverseIce mentioned on Twitter that this score is lower than the Snapdragon 865, I'd like to mention that you can not under any circumstances compare Antutu scores from the Android app with the Antutu scores from the iOS app.

I always have a good laugh if people say that Snapdragon chips are faster than the Apple A series chips (or the other way around) and then post an Antutu benchmark.

You can read this Antutu blogpost that explain why they're not comparable here: http://www.antutu.com/en/doc/119646.htm

What about Geekbench? ? ? ?
At least antutu is somehow open source, but geekbench is not. That says a lot about theyr code... ? It's so funny people still believe Bionic being faster than an I7 or ryzen based on geekbench software which is NOT open source! ???

 

[NotTooLate]

What about Geekbench? ? ? ?
At least antutu is somehow open source, but geekbench is not. That says a lot about theyr code... ? It's so funny people still believe Bionic being faster than an I7 or ryzen based on geekbench software which is NOT open source! ???

No need to guess anymore , we are soon going to compare apples to apples , same SW doing the same work.

 

[supertiffany]

the iPhone 11 has the fastest chip by a mile among the competition

would rather have iPhone 12 chip have significant battery efficiency to make up to that 10 percent smaller battery size

tho it was said in leaks that it's 30 percent more battery efficient

 

[Motionblurrr]

Okay, so I'm curious..

Was the iPhone 11/11PRO generation considered an S year? Or was that (obviously by the name) the XS generation? Even though the hardware looks different it would be the 11/11Pro's S version in a sense right?

Because the XS generation is still crazy powerful. I'm betting on Apple focused on software and the 30% more efficiency (even though it might come with a smaller battery for example) will net them a 2hour growth in battery life or more.

Meanwhile, the top of the line Samsung Note 20 Ultra are getting measly average 5 hours SOT with 120hz enabled. Which makes sense on why Apple is taking their sweet time with 120hz.

 

[The Game 161]

Okay, so I'm curious..

Was the iPhone 11/11PRO generation considered an S year? Or was that (obviously by the name) the XS generation? Even though the hardware looks different it would be the 11/11Pro's S version in a sense right?

Because the XS generation is still crazy powerful. I'm betting on Apple focused on software and the 30% more efficiency (even though it might come with a smaller battery for example) will net them a 2hour growth in battery life or more.

Meanwhile, the top of the line Samsung Note 20 Ultra are getting measly average 5 hours SOT with 120hz enabled. Which makes sense on why Apple is taking their sweet time with 120hz.

You do realise android and ios measure screen on time differently, right? So it's not that simple

Exynos versions are at 5 hours but SD is anywhere from 6 to 9 hours screen time

Battery life is not why 120hz didn't make it... It was more display issues

 

[EugW]

I wonder if this date means anything. From the iPad event:

 

[mprime_17]

the iPhone 11 has the fastest chip by a mile among the competition

would rather have iPhone 12 chip have significant battery efficiency to make up to that 10 percent smaller battery size

tho it was said in leaks that it's 30 percent more battery efficient

There were some rumors saying that Mediatek (?) was getting 40% power reduction with TSMC's 5nm. So that 30% better efficiency for apple doesn't sound too off. I'd also prefer that they focus on better battery instead of more performance.

 

[snipr125]

With phones what matters the most is single core performance, and in this regard Android flagships with the SD 865 have not even caught up to the single core performance of the A11 in the iPhone 8/8 plus/X from 2017. E.g OnePlus 8 with 900 on Geekbench 5 with the iPhone 8/X on 916!!!

 

[EugW]

With phones what matters the most is single core performance

No.

 

[snipr125]

No.

I was giving my opinion regarding single core being more important and not stating a fact, but do you mind elaborating further why you think No?

 

[EugW]

I was giving my opinion regarding single core being more important and not stating a fact, but do you mind elaborating further why you think No?

First off, Apple non-X SoCs have been 4 core or more for years.

Some activities are better multi-threaded than others, but there is always still a level of multi-tasking happening in these phones. Phones in 2020 don't just run one service/app/tab at a time. However, in general these days most of the activities which actually need true fast performance are usually pretty well multi-threaded. For example, a banking app doesn't use much multi-threading, but it doesn't matter since even older SoCs have single-core speed that's already fast for this. You're not going to notice any difference between an A10 and A12 for a banking app, despite the fact that A12 is much faster than A10 for single-core (and multi-core). This is also generally true for accessing mobile versions of web sites on an iPhone. Yeah, they tend to be more single-core limited, but even on A10 they already run fast. In contrast, if you're trying to do a software video decode on a third party app, performance is key, and this is well multi-threaded. And if you're doing hardware video decoding, neither single-core nor multi-core CPU performance is really important here, since it's being done on purpose-built dedicated silicon whether it's on the GPU or a separate chip or whatever.

 

[snipr125]

First off, Apple non-X SoCs have been 4 core or more for years.

Some activities are better multi-threaded than others, but there is always still a level of multi-tasking happening in these phones. Phones in 2020 don't just run one service/app/tab at a time. However, in general these days most of the activities which actually need true fast performance are usually pretty well multi-threaded. For example, a banking app doesn't use much multi-threading, but it doesn't matter since even older SoCs have single-core speed that's already fast for this. You're not going to notice any difference between an A10 and A12 for a banking app, despite the fact that A12 is much faster than A10 for single-core (and multi-core). This is also generally true for accessing mobile versions of web sites on an iPhone. Yeah, they tend to be more single-core limited, but even on A10 they already run fast. In contrast, if you're trying to do a software video decode on a third party app, performance is key, and this is well multi-threaded. And if you're doing hardware video decoding, neither single-core nor multi-core CPU performance is really important here, since it's being done on purpose-built dedicated silicon whether it's on the GPU or a separate chip or whatever.

Ok thanks for responding EugW. Yes i definitely agree with you regarding intense tasks like video encoding/music production/batch photo editing requiring more cores/multi threading, but wouldn't you agree that the vast majority of basic daily tasks on an iphone mainly rely on single core performance for efficiency, speed and battery life?

 

[EugW]

Ok thanks for responding EugW. Yes i definitely agree with you regarding intense tasks like video encoding/music production/batch photo editing requiring more cores/multi threading, but wouldn't you agree that the vast majority of basic daily tasks on an iphone mainly rely on single core performance for efficiency, speed and battery life?

I guess the best argument there is battery life. My point though was perceived speed isn’t an issue anymore for most such activities. A12 is already very, very fast in that regard. That’s why on iPhone SoCs, Apple prioritized small cores not big. There are four efficiency cores and two performance cores.

Put it this way, for such basic daily activities as you descrribe, I don’t feel that even the A10 on the 7 Plus is slow on an iPhone, and that’s a four year old SoC. However, I can notice occasional pauses, which I attribute to insufficient RAM, not SoC performance.

It’s different on an iPad though. Whereas A10 is very fast with mobile websites aimed at phones for example, you now can feel some slowdowns on the desktop sites you get on iPads. It’s not bad at all on A10, but the difference can be noticeable, esp. when you multi-task. Interestingly though, multi-core largely solves this problem too, as an A10X iPad Pro feels faster for surfing desktop sites than A10 does.

—-

Meanwhile, some Twitter user is claiming the iPhone 12 mini will have a lower performance chip called B14, postulated to be a downclocked version of A14.

Apple will reportedly use a scaled-down version of the A14 Bionic on the iPhone 12 Mini

The soon-to-be-released iPhone 12 Mini will allegedly run a scaled-down version of the A14 Bionic. Dubbed as the Apple B14, the chip will also appear alongside subsequent variants of the iPhone SE. The B14 is, in all likelihood, an underclocked A14 Bionic fine-tuned for efficiency instead of...

www.notebookcheck.net

https://twitter.com/i/web/status/1305699024947875842

 

[snipr125]

I guess the best argument there is battery life. My point though was perceived speed isn’t an issue anymore for most such activities. A12 is already very, very fast in that regard. That’s why on iPhone SoCs, Apple prioritized small cores not big. There are four efficiency cores and two performance cores.

Put it this way, for such basic daily activities as you descrribe, I don’t feel that even the A10 on the 7 Plus is slow on an iPhone, and that’s a four year old SoC. However, I can notice occasional pauses, which I attribute to insufficient RAM, not SoC performance.

It’s different on an iPad though. Whereas A10 is very fast with mobile websites aimed at phones for example, you now can feel some slowdowns on the desktop sites you get on iPads. It’s not bad at all on A10, but the difference can be noticeable, esp. when you multi-task. Interestingly though, multi-core largely solves this problem too, as an A10X iPad Pro feels faster for surfing desktop sites than A10 does.

—-

Meanwhile, some Twitter user is claiming the iPhone 12 mini will have a lower performance chip called B14, postulated to be a downclocked version of A14.

Apple will reportedly use a scaled-down version of the A14 Bionic on the iPhone 12 Mini

The soon-to-be-released iPhone 12 Mini will allegedly run a scaled-down version of the A14 Bionic. Dubbed as the Apple B14, the chip will also appear alongside subsequent variants of the iPhone SE. The B14 is, in all likelihood, an underclocked A14 Bionic fine-tuned for efficiency instead of...

www.notebookcheck.net

https://twitter.com/i/web/status/1305699024947875842

Yes multi core on ipads is more important, as one is more likely to do intensive tasks with the larger screen, pen and ability to attach keyboard etc. A14 on the new ipad air 4 is a very exciting prospect in this regard. Now this tweet about a B14 is very interesting, but if the iphone 12 5”4 gets this chip then sorry but im not getting it. My iphone SE already has an A13 so why would i downgrade to a mid range SoC? Maybe this is Apples way of upselling, which they are very good at as we all know.

 

[EugW]

Yes multi core on ipads is more important, as one is more likely to do intensive tasks with the larger screen, pen and ability to attach keyboard etc. A14 on the new ipad air 4 is a very exciting prospect in this regard. Now this tweet about a B14 is very interesting, but if the iphone 12 5”4 gets this chip then sorry but im not getting it. My iphone SE already has an A13 so why would i downgrade to a mid range SoC? Maybe this is Apples way of upselling, which they are very good at as we all know.

I've mentioned this before elsewhere but I'll mention it here too: Apple's main target market for new iPhones is not the previous year's iPhone owners. It's everyone else. If you already have an A13 iPhone, Apple already got your money recently, so effectively you're off their radar.

It should be noted that the average upgrade cycle for iPhones now is 3 to 4 years, depending upon which analyst and which dataset you believe. And remember, that is average, not longest. That means there are a bunch of people waiting 5 years before upgrading. These people with 3-4-5 year-old phones are what Apple is targeting.

BTW, as mentioned, I am currently using an iPhone 7 Plus, with A10 which is 4 years old. My wife has the XR with A12, but it effectively feels about the same speed with basic iPhone apps. Maybe the XR has a slight edge, but it's nothing dramatic. (Both have 3 GB RAM.) I suspect most mainstream users wouldn't even notice the difference (aside from Face ID, etc.) until they started gaming or something.

I am strongly considering upgrading this year, but it's not because of the A14 chip. It's because of the camera upgrades, 6 GB RAM, and 5G, in that order. OLED will also be a bonus, although I am a bit worried about possible PWM headaches. When it comes to these refresh rate frequency related types of things, I have generally been a bit more sensitive than average, but not overly sensitive, so we'll see how that goes.

P.S. I am very skeptical of that B14 Twitter post, but I thought I'd post it anyway.

 

[Tsepz]

I guess the best argument there is battery life. My point though was perceived speed isn’t an issue anymore for most such activities. A12 is already very, very fast in that regard. That’s why on iPhone SoCs, Apple prioritized small cores not big. There are four efficiency cores and two performance cores.

Put it this way, for such basic daily activities as you descrribe, I don’t feel that even the A10 on the 7 Plus is slow on an iPhone, and that’s a four year old SoC. However, I can notice occasional pauses, which I attribute to insufficient RAM, not SoC performance.

It’s different on an iPad though. Whereas A10 is very fast with mobile websites aimed at phones for example, you now can feel some slowdowns on the desktop sites you get on iPads. It’s not bad at all on A10, but the difference can be noticeable, esp. when you multi-task. Interestingly though, multi-core largely solves this problem too, as an A10X iPad Pro feels faster for surfing desktop sites than A10 does.

—-

Meanwhile, some Twitter user is claiming the iPhone 12 mini will have a lower performance chip called B14, postulated to be a downclocked version of A14.

Apple will reportedly use a scaled-down version of the A14 Bionic on the iPhone 12 Mini

The soon-to-be-released iPhone 12 Mini will allegedly run a scaled-down version of the A14 Bionic. Dubbed as the Apple B14, the chip will also appear alongside subsequent variants of the iPhone SE. The B14 is, in all likelihood, an underclocked A14 Bionic fine-tuned for efficiency instead of...

www.notebookcheck.net

https://twitter.com/i/web/status/1305699024947875842

Wow a very interesting rumor that, and it does make sense, will be interesting to see how efficient they will make that B14 versus chips like the Snapdragon 765G, while keeping it powerful.