futuremark Does Apple throttle performance for old iPhones? Futuremark settles the question

[ScratchCat]

Just now, DrMacintosh said:

Doesn’t really matter if their parallelism can’t even compete with the A11 on any level. 

True, I'm not to sure what android SoC manufacturers are thinking right now, the A53 has been replaced with a 10% faster A55 which is less power efficient , the Kyros in the 835 are lower in some aspects ie integer than the 821.

 

Please not that this normally is in geekbench, and I really doubt the ~6W 2.3 GHz A11 has a similar single thread performance than my i5 4690k at 3.9 GHz and ~50W.

[DrMacintosh]

8 minutes ago, ScratchCat said:

and I really doubt the ~6W 2.3 GHz A11 has a similar single thread performance than my i5 4690k at 3.9 GHz and ~50W.

If I’m not mistaken the Mobile and Desktop versions run different tests. 

 

Of course they aren’t as powerful. If they were Apple would be running x86 CPU manufactures out of business since Intel and AMD clearly incompetent  

[sinkesnnj]

People tend to forget about 3rd party impact. I am on Android and remember when FB apk was 50mb total. Nowadays when you install FB it's a 350MB apk. Of course it won't run as smooth as it once did on hardware that barely ran it when it was 50MB. Not to mention every new iteration of OS adds new features that drain more resources from a device. 2017 apk and OS are built for 2017 hardware. Anyone expecting it to run smoothly on 2013 hardware, with 50% less RAM, 50% less battery and 40% slower SOC is an idiot...

[bcredeur97]

I never doubted the phones outright performance degrading... what I have noticed personally over the years:

1. More complex animations; loading the GPU more (Though this is less now...)

2. More processes running at the same time on newer versions of iOS

3. More memory being used. This is a big one. Every iOS update seems to use about 100-200MB more memory. This has also been less lately though. 
Just going from iOS 3 -> iOS 6 was like going from your phone using 100MB or so of RAM while idle to nearly 500MB. It was a big thing back then when the phones only had 256 or 512MB of RAM.

 

really processing power hasn't really been a problem for me since the iPhone 6ish. The biggest problem I see now is memory... If they crammed 4-6GB of RAM into phones even though it's ridiculous right now (on iOS anyway...) then that would go a LOOOOONG way

[ScratchCat]

4 minutes ago, DrMacintosh said:

If I’m not mistaken the Mobile and Desktop versions run different tests. 

 

Of course they aren’t. If they were Apple would be running x86 CPU manufactures out of business since Intel and AMD clearly incompetent  

They then really need to update their website :

Quote

Designed from the ground-up for cross-platform comparisons, Geekbench 4 allows you to compare system performance across devices, processor architectures, and operating systems. Geekbench 4 supports Android, iOS, macOS, Windows, and Linux

 

 

3 minutes ago, bcredeur97 said:

I never doubted the phones outright performance degrading... what I have noticed personally over the years:

1. More complex animations; loading the GPU more (Though this is less now...)

2. More processes running at the same time on newer versions of iOS

3. More memory being used. This is a big one. Every iOS update seems to use about 100-200MB more memory. This has also been less lately though. 
Just going from iOS 3 -> iOS 6 was like going from your phone using 100MB or so of RAM while idle to nearly 500MB. It was a big thing back then when the phones only had 256 or 512MB of RAM.

These are the points which matter, an iPhone 4s can hardly open 2 tabs without running out of RAM today and we didn't always have to deal with Siri trying to be useful and recommending the most useless items. When am I ever going to ask " will it be hot today? " at 11pm?

[Mr.Meerkat]

2 hours ago, Hayabusa1989 said:

Im not really making assumptions as Apple had camera inserts in the IPAD 1 but did not put a camera in ... its the sorta thing they like to do. 

And? Don't forget how people who uses tablet to take pictures get frowned upon  

 

Anyway, even if they added the camera(s) in, it would have probably been the same camera as the iPod touch 4th gen (spoiler alert, the iPad 2 was indeed equipped with em) and as an owner of a iPod touch 4th gen, lets just say that you shouldn't really be taking photos with em  

[vorticalbox]

5 hours ago, hey_yo_ said:

I think Apple should at least add more RAM to their iPhones and iPads for future proofing.

why would they want to future proof? That means you take longer to buy a new iPhone. 

[DrMacintosh]

3 minutes ago, vorticalbox said:

That means you take longer to buy a new iPhone. 

That’s flawed. iPhone already last about 5 years.......

 

Them getting slow is not why people buy new iPhones. 

[tsk]

1 hour ago, DrMacintosh said:

In all these cases is pretty much due to animations. iOS has become very “dainty” in that regard. It can still beat the 835 when the time for lifting weights comes around. Remember the A11 is a hex core with 4 high efficiency cores and 2 High Power cores. The A11 is using its high efficiency cores for pretty much every app that isn’t rendering. 

 

The Apple A11 beats the SnapDragon 835 handedly. In Geekbench well over 4,000 points. 

 

 

It's not just down to animation. The A11 beats the S835 in pretty much every benchmark as well as most in-app performance. Android beats iOS in OS performance, that's what I'm talking about, and those are the slowdowns people experience with iOS upgrades. 

[ScratchCat]

18 minutes ago, vorticalbox said:

why would they want to future proof? That means you take longer to buy a new iPhone. 

Lots of people buy new phones because they either break or get bored of their old phone. Additionally you have the younger crowd who buy the newest phone because it is 'cool', 'hip' and all the other things which require google translate from youth -> english (jk).

I expect a lot of purchases of the iPhone X are due to wanting a phone which looks and feels different.

[79wjd]

3 hours ago, ScratchCat said:

I expect a lot of purchases of the iPhone X are due to wanting a phone which looks and feels different.

I can't speak for others, but there was always something that compelled me to upgrade. Although I certainly did want to upgrade more frequently since my Razr, but I just couldn't rationalize spending the money -- plus it used to be "harder" since everything was tied to contracts.

• Droid X (first smartphone) -- kept for about 7 months
• HTC Thunderbolt (I wanted 4G, although damn was that a shitty phone) -- kept for about 13 months.
• Razr Maxx (Thunderbolt's battery life was abysmal and it got slow) -- kept for about 18 months.
• iPhone 5s (my Razr became painfully slow so I overclocked it to solve that, but then I didn't receive half my messages and it still had problems -- the screen looked awful and the camera was terrible) -- kept 2 years.
• iPhone 6s+ (wanted a bigger screen, my 5s still works pretty well even today on iOS 11) -- kept 2+ years.
• iPhone X (I want a smaller phone without sacrificing the screen, plus the better camera(s) is nice too).

[vanished]

who decided this was an intelligent way to display this data?

 

 

Other than the obvious "make the bars taller and fill the empty space", they could have used a cut axis to really zoom in on the differences, if any.  At this scale they all looks basically the same.

[captain_to_fire]

9 minutes ago, Ryan_Vickers said:

who decided this was an intelligent way to display this data?

 

Other than the obvious "make the bars taller and fill the empty space", they could have used a cut axis to really zoom in on the differences, if any.  At this scale they all looks basically the same.

I think they did the bar graph with Microsoft Excel. I think the guys at Futuremark forgot that they can reduce the empty space within Excel. 

[vanished]

19 minutes ago, hey_yo_ said:

I think they did the bar graph with Microsoft Excel. I think the guys at Futuremark forgot that they can reduce the empty space within Excel. 

Anyone who's been through elementary school knows how to make a chart, if a website that specializes in benchmarking can't figure it out they have some serious explaining to do, and some serious problems to rectify.

[Belgarathian]

10 hours ago, Not_Jony_Ive said:

The real question is: Do android updates slow down older android devices?? 

 

  Reveal hidden contents

/s

Wouldn't have enough updates to test

[Commodus]

I won't rehash the earlier arguments, but I will say this: Apple is getting better about maintaining UI performance over time.

 

Some of it comes simply from speed improvements.  An iPhone 5S is much faster than an iPhone 4, so it's not going to struggle as much when you throw new features at it.  But more importantly, Apple changed its approach to how it supports older devices.  It used to be that it developed for recent hardware and just figured out which features it had to pull to make the OS run on older hardware.  Now, it's more holistic, taking those devices into account earlier into the process.  This doesn't mean that your 4-year-old phone will run iOS as well as the device released last week, but it could be the difference between "holy hell, this is rough" and "it's slower, but tolerable."

 

As it is, I'd rather take Apple's long-term support approach, where UI performance can get iffy on older hardware, than what you tend to get with Android OEMs.  Yeah, Google Play Services and other updates can keep a device running apps and newer features longer than the OS version would suggest, but you still end up in a situation where your device is dropped like a hot potato: you get two years of major OS feature updates if you're lucky, and security updates rarely last much longer (if at all).  There's also a kind of class system where you're frequently 'punished' for buying a lower-end phone by getting less support.  I've seen some Android phones where the OS version you get in the box is the one you're stuck with forever.

[Hayabusa1989]

16 hours ago, Mr.Meerkat said:

And? Don't forget how people who uses tablet to take pictures get frowned upon  

 

Anyway, even if they added the camera(s) in, it would have probably been the same camera as the iPod touch 4th gen (spoiler alert, the iPad 2 was indeed equipped with em) and as an owner of a iPod touch 4th gen, lets just say that you shouldn't really be taking photos with em  

hmm this does not dispute anything i said though... 

[Commodus]

3 hours ago, huilun02 said:

Pfft slow down. Cause? Fixes. LMAO

Android phones don't slow down. They can speed up. How? XDA.

 

Do you even know how to phone bruh? 

I never like using "just root it and install X" as an answer to phone problems.  It's basically an admission that you're willing to settle for bad practices out of the box because hey, you can fix it later.  Like going to a restaurant with mediocre food because you can 'fix' it with sauce you brought from home.  Yeah, you took care of it, but you've also told the provider that you're perfectly willing to pay for their crap.  They never get the hint that they need to change.

[captain_to_fire]

22 minutes ago, Commodus said:

I never like using "just root it and install X" as an answer to phone problems.  It's basically an admission that you're willing to settle for bad practices out of the box because hey, you can fix it later.  Like going to a restaurant with mediocre food because you can 'fix' it with sauce you brought from home.  Yeah, you took care of it, but you've also told the provider that you're perfectly willing to pay for their crap.  They never get the hint that they need to change.

Indeed. In fact, corporations would prevent a rooted/jailbroken phone to connect to company resources because it is running tampered software. Some companies allow BYOD but their IT personnel have to check the phones for any signs of a jailbreak or root. No wonder my previous employer stopped issuing Android phones to employees because Samsung in particular is slow in deploying security updates. Now the only phone they allow are iPhones.

[captain_to_fire]

2 minutes ago, huilun02 said:

IT personnel can check my phone all they want and after they're done my Oneplus device will have its bootloader unlocked and Magisk installed in less than 5 minutes.

There's a reason why many companies have a strict policy against rooted Android phones and why they ban them.

8 minutes ago, huilun02 said:

This is just me taking actual ownership of my device, defying all them manufacturers' greedy ideas of how I should use my phone or when I should buy a new phone.

I can understand where people rooting their phones are coming from. I think there's a ROM that allows a GS4 to run Nougat or any other newer Android version. But then for a lot of people especially non-techies, rooting their phones themselves is something not feasible.

[Mr.Meerkat]

8 hours ago, Hayabusa1989 said:

hmm this does not dispute anything i said though... 

I know, just pointing out the fact that cameras on tablets are pointless (unless we're talking about having a front facing camera, which is nice to have for video chats). 

[kokakolia]

On 10/10/2017 at 5:24 AM, Teddy07 said:

That is true but I have to spend money and why should I when my S2 does the job? I don't use my phone that much

I know exactly where you're coming from. If you have a Galaxy S2 that's never ever been updated it runs so fast and buttery smooth. Same can be said about the iPhone 4. I almost believe that these were the best versions of Android and iOS. It's shocking to see something run so well on such puny hardware. 

[suits]

I don’t know about performance. But my battery sure has tanked since iOS 11

[Fgtfv567]

I think there's a serious disconnect between an SoC's raw performance versus the Real World User Experience of a phone.

 

OK, the raw performance of the phone under benchmarks didn't change, and that's still good, but benchmarks are not the way to measure a phone's smoothness (or how much it stutters).

I wish the researcher would have contacted one of the writers at XDA-Developers to learn more about real world UX and tools to investigate how smooth a phone really is, such as GPU Profiling

[kokakolia]

9 minutes ago, Fgtfv567 said:

I think there's a serious disconnect between an SoC's raw performance versus the Real World User Experience of a phone.

 

OK, the raw performance of the phone under benchmarks didn't change, and that's still good, but benchmarks are not the way to measure a phone's smoothness (or how much it stutters).

I wish the researcher would have contacted one of the writers at XDA-Developers to learn more about real world UX and tools to investigate how smooth a phone really is, such as GPU Profiling

GPU profiling? 

 

How about timing:

 

- booting the phone?

- launching a YouTube video?

- loading a picture heavy site like CNN?

- taking a photo?

- sending a text?

- calling someone from a contacts list?

 

I am talking real world activities!