Benchmarks suggest Apple doesn't slow down old iPhones with each new iOS release, but...

[midian182]

You’ve probably heard the theory that every new software update put out by Apple causes its older iPhones to slow down. It’s a common belief, but benchmarking firm Futuremark has spent the last year trying to debunk the theory.

Apple isn’t the only firm accused of using “planned obsolescence” as a way of getting consumers to upgrade. Rumors that Nvidia hampers the performance of its older GPUs through driver updates have been around for years, though YouTube channel Linus Tech Tips recently proved this wasn’t the case.

Futuremark, the company behind the likes of PCMark10, began collecting iPhone performance data in 2016. Using its own 3DMark free benchmarking app, it conducted over 100,000 tests for seven different phone models across three versions of iOS — 9, 10, and 11.

Starting with the 2013 iPhone 5s, the tests show that the handset’s GPU and CPU performance has remained consistent throughout each iOS update. It’s the same story with the iPhone 6, 6s, and 7, all of which show only slight CPU drops, which Futuremark claims users would be unlikely to notice. It attributed these declines to minor iOS updates and other factors.

The benchmarking company says the results prove Apple doesn’t intentionally slow down the performance of older iPhones with each new iOS release. It does, however, concede that updates might add new features that use more resources or require more processing power, which might affect "people's perception of performance."

But the biggest problem with these tests is that they don’t tell the whole story. Many commentators point out that they’re not an accurate reflection of real-world performance. There are numerous reports of GUI lag following an iOS update, while poor keyboard response times are also a problem. Apple might not be purposely hindering older iPhones to force users to upgrade, but it appears that optimizing each new version of iOS to work with these devices isn’t the company's highest priority.

Permalink to story.

https://www.techspot.com/news/71309-benchmarks-suggest-apple-doesnt-purposely-slow-down-old.html

 

[yRaz]

Can't wait for the people with the $1000 iPhone x to have these slowdowns

 

[oldikes]

I dont so much experience lower performance as much as I experience slight hangs or longer startup/switching times for apps and stuff, which I dont believe benchmarks would capture so much. but I am on a 5C which by modern standards is a steaming pile of slowness.
the fact that iphone 5S and ipad air 1 support iOS11 is fantastic. These devices are 4 years old.

 

[Skidmarksdeluxe]

Can't wait for the people with the $1000 iPhone x to have these slowdowns

You'll have to wait until next year time but by then the $1000 price tag won't be an issue because they won't be worth $100.

 

[Rockstarrrr]

What these companies do is not actually slowing down - its lack of certain optimization. iPhones do work slower with newer iOS versions and nVidia GPUs do badly in newer titles simply because nVidia couldnt care less about performance on GTX 650-Ti or 770. What is example of optimizations? Remember when Mirrors Edge Catalyst was using more VRAM than 3.5 GB GTX 970 could handle? They used "trick" to prevent game to use defective 512 MB VRAM by basically reverting from "hyper" settings without user authority (in order to give you good performance). Yes, Keplers and Fermies lack certain hardware but they could still be usable as some also newer titles showed.

Then you have some where 1050 is easily faster than 780-Ti.

 

[FF222]

The benchmark only shows that Apple doesn't lower the clock rate of the iPhones with the updates. However, it does not measure their impact on UX speed, memory consumption, etc., which all would affect the actual speed the apps are launching and running with on the devices.

 

[Puiu]

Synthetic tests =/= OS and app speeds. The hardware stays the same after all. Planned obsolescence is not something imaginary.

 

[Demur]

After updating my iPhone 6 Plus to iOS 11, it absolutely 100% got slower. I only really use a few apps like Safari, Google Music, Youtube and Discord, but all of them started chugging after I updated. They're still not performing well after subsequent updates, both to iOS and the apps themselves.

I don't know whether or not it's actually deliberate and I don't know what specific element is causing this problem, but Apple absolutely screwed something up. I certainly don't appreciate updates that add a handful of features I'll never use in exchange for worse performance, and I think it's perfectly reasonable for people to interpret it as a not-so-subtle nudge from Apple to upgrade to newer phones.

 

[fadingfool]

Not sure if these are the correct benches to show what they wanted to show (the trend lines for most of them do have a slight negative gradient - lack of scale doesn't help either). Responsiveness and app loading/switching/closing times are the main interface that people notice not benchmarks showing top CPU or GPU processing capabilities. This seems to be a pro-Apple PR exercise using an inappropriate metric to "disprove" what a lot of people are experiencing. Now if they did a few different tests, latency metrics and app metrics then I would be more convinced.

Edited to add - no-one accused apple of lowering clock rates so it seems this is a case of "Apple haven't done what you haven't accused them of doing."

 

[Slavjan]

These benchmarks are about as good as showing average FPS without the minimums.

 

[Theinsanegamer]

What these companies do is not actually slowing down - its lack of certain optimization. iPhones do work slower with newer iOS versions and nVidia GPUs do badly in newer titles simply because nVidia couldnt care less about performance on GTX 650-Ti or 770. What is example of optimizations? Remember when Mirrors Edge Catalyst was using more VRAM than 3.5 GB GTX 970 could handle? They used "trick" to prevent game to use defective 512 MB VRAM by basically reverting from "hyper" settings without user authority (in order to give you good performance). Yes, Keplers and Fermies lack certain hardware but they could still be usable as some also newer titles showed.

Then you have some where 1050 is easily faster than 780-Ti.

Going to correct a few things here.

First of all, please show the benchmarks of a 1050 beating a 780ti. First you guys were saying the 1050ti could beat it, now its a 1050? Benchmarks consistently show it to be around 1060 level, which is where you would expect a 2 generation old flagship to be, right around mid range.

Second, the reason iOS devices work slower is not due to lack of optimization. Did you not RTFA? If apple was not optimizing, then GPU/CPU performance would have slowed down in newer versions of iOS. It doesnt.

The reason apple devices slow down is the same reason old android phones do: mobile hardware is still evolving at a very rapid pace. Compare a current snapdragon 820, now a year old, to the three year old 800 or two year old 810. the performance jumps were huge between generations!. Memory speed and storage speed are making similar jumps.

5 years in the mobile world is more similar to 10-15 years in the desktop world. The iphone 5s was a good device, but it is very old today. Nobody expects them to run like new, just like nobody expected pentum IV chips to run windows 7 properly with no slowdowns.

Why people think that phone OSes will not grow to be more demanding like desktop OSes do is beyond me. Seems a lot of people want to have their cake and eat it to when it comes to phones.

 

[mattfrompa]

Can't wait for the people with the $1000 iPhone x to have these slowdowns

You'll have to wait until next year time but by then the $1000 price tag won't be an issue because they won't be worth $100.

Closer to $100 off

 

[Skidmarksdeluxe]

Closer to $100 off

Closer to $100 off... the $100 the seller was expecting.
I just completed what you were trying to say before you were distracted by something.

 

[thelatestmodel]

" it appears that optimizing each new version of iOS to work with these devices isn’t the company's highest priority."

... which is planned obsolescence. "It's slow now, guess I'll buy a new one."

If your software update slows down older devices because of new features, make those features optional on older devices or leave them out altogether. Advise customers what the update is going to do, and what performance will be like.

But no, this is Apple's world, what am I talking about.

 

[Erik Vianna]

I believe it's not about the iOS itself, what about the apps? I'm pretty sure the apps got a lot more demanding over time and devices with weaker hardware gets hit harder.

Same thing with Android, my old Xperia running Gingerbread, the Google Play Services would take up to 20mb of RAM (it was a single core Snapdragon 1.4Ghz 512mb ram) but the phone was superb until stuff started updating. Just looking at my Galaxy S7e the amount of RAM Google apps alone take is hitting around the 200MB+ mark.

Same stuff with a friend's phone with 2GB ram that used to take whatever you throw at it, things are slowly getting terribly slow. After installing a AOPK Rom it kinda gave it some extra life.

People don't know how to use their devices, I see iPhone 5s users with Facebook, and four different messengers apps running in background plus all kind of ****. And yes sure let's leave WiFi, Bluetooth, Location services on because it shouldn't drain as much battery from a four years old phone. Oh sure Apple is to blame.

 

[mattfrompa]

Closer to $100 off... the $100 the seller was expecting.
I just completed what you were trying to say before you were distracted by something.

Nope I mean last years model would be sold to $100 off. I don't know what context would make it worth $100