Geekbench Chief: Android Stagnates While iPhone Soars

[Mark Spoonauer]

The founder of Primate Labs is sounding off on the performance gap between the iPhone 8 and Android flagships, and what it means for mobile computing.

Geekbench Chief: Android Stagnates While iPhone Soars : Read more

 

[rogo.jack17]

Okay everybody has figured out how to manipulate the performance test on these graphics sites and also by writing inline diss or aol or iOS programs to create a fault score on the index. what makes you think that somebody couldn't write a program insert it into the iOS that manipulates the score has anybody looked into that. okay I've had an iPhone 8 and an lgv30 side-by-side with processing the same graphics and processing the same screens as far as loading and unloading the blocks and they're pretty much exactly the same the iPhone beat it on loading the graphics that 3D rotation and things like that the v30 beat the Apple so not really sure where you're getting your information but it could be skewed I physically held them in my hand and done that have you?

 

[Mark Spoonauer]

Your point is taken on synthetic benchmarks but not only have we hold the phones in our hands we've compared their performance in real-world tests. This includes video editing, opening large files and opening larger apps. The iPhone 8 came out on top in each case.

https/www.tomsguide.com/us/iphone-8-benchmarks-fastest-phone,review-4676.html

Okay everybody has figured out how to manipulate the performance test on these graphics sites and also by writing inline diss or aol or iOS programs to create a fault score on the index. what makes you think that somebody couldn't write a program insert it into the iOS that manipulates the score has anybody looked into that. okay I've had an iPhone 8 and an lgv30 side-by-side with processing the same graphics and processing the same screens as far as loading and unloading the blocks and they're pretty much exactly the same the iPhone beat it on loading the graphics that 3D rotation and things like that the v30 beat the Apple so not really sure where you're getting your information but it could be skewed I physically held them in my hand and done that have you?

 

[adamcawthorn]

Has anyone considered that the operating systems are different and if you swapped the chips over to handle each other's OS that things may be different? Just a thought. I don't know enough to know if this could be a factor.

 

[Mark Spoonauer]

Interesting thought. Apple has long touted that it has an advantage because it controls both the hardware and software and how they're integrated. Google buying HTC could help Android,
at least with Pixel phones.

Has anyone considered that the operating systems are different and if you swapped the chips over to handle each other's OS that things may be different? Just a thought. I don't know enough to know if this could be a factor.

 

[Phil_29]

Mark, I have to disagree with you. First of all android may always lag behind because its sold to the various vendors and they change it as needed. Its not in house so to speak. The key word there is "may." So your taking a google designed operating system, modified by Samsung to their needs and for their processor. Along the way some "optimization is changed. Its inevitable.
My girlfriend has an 8 and I have the note 8. They are neck and neck in everything.
The other thing to consider is this.. for your average user that does NOT do video editing, or opens super huge files.. will not likely notice any difference between the two handsets.
There are things that so called power users to..like test against video editing.. I have never met anyone who does video editing on a phone.
Of all the people I know, text, facebook, pictures, games, internet, open apps and check email. SO why not compare things that people actually do on a daily basis.

So lets compare things people actually DO on their phones. Its like saying a corvette will blow away a honda accord on internate 95. Ok because the speed limit on interstate 95 is what, 55-65?
How many people drive a corvette? Not many. Honda accord, a shit load. If you want to do a fair comparison between cards, take a car that people drive and use daily like a nissan vs a honda.

How many people do video editing? not many. How many people use the internet, text and open small aps.. a lot..
And do keep in mind that optimization on apples part probably does have an edge because its in house.

 

[Mark Spoonauer]

Thanks. Video editing is only one application. The Note 8 also fell behind on opening a large file, something a lot of people would do as an attachment, and in opening a game (though less dramatic). But as noted, it's also about what's coming next. AR apps in particular take up a lot of resources.

We will be following up with other everyday tasks as well.

Mark, I have to disagree with you. First of all android may always lag behind because its sold to the various vendors and they change it as needed. Its not in house so to speak. The key word there is "may." So your taking a google designed operating system, modified by Samsung to their needs and for their processor. Along the way some "optimization is changed. Its inevitable.
My girlfriend has an 8 and I have the note 8. They are neck and neck in everything.
The other thing to consider is this.. for your average user that does NOT do video editing, or opens super huge files.. will not likely notice any difference between the two handsets.
There are things that so called power users to..like test against video editing.. I have never met anyone who does video editing on a phone.
Of all the people I know, text, facebook, pictures, games, internet, open apps and check email. SO why not compare things that people actually do on a daily basis.

So lets compare things people actually DO on their phones. Its like saying a corvette will blow away a honda accord on internate 95. Ok because the speed limit on interstate 95 is what, 55-65?
How many people drive a corvette? Not many. Honda accord, a shit load. If you want to do a fair comparison between cards, take a car that people drive and use daily like a nissan vs a honda.

How many people do video editing? not many. How many people use the internet, text and open small aps.. a lot..
And do keep in mind that optimization on apples part probably does have an edge because its in house.

 

[Phil_29]

and by the way, no disrespect meant, just a friendly debate

 

[Phil_29]

ok so here's a question. If apple is so superior, why does android hold most of the market share?
If you are going to talk about speed, the bionic may be a bit faster than the samsung, however, feature wise, the note 8 blows the i phones out of the water.
for example, wireless charging, ive had for a while and NOW apple is coming out with it?
They may be a bit faster in processing, that I highly doubt your average user will notice, but where are they when it comes to things samsung has already had in the market?

Is the bionic the only thing that apple offers that may be a bit faster? if apple was superior, and I do like apple stuff, honestly, they would dominate the market place and they simply are not.
Great article, although it reminds me of things that I read being a pc power user that claim that the geforce a gives you 5 fps more than the geforce b.. because the human eye can only notice so much and 5 fps is not noticeable to your average person..

 

[Mark Spoonauer]

Of course and thank you. We need to devise more real-world performance tests to help shoppers decide what's best for them. And Qualcomm has a new chip on the horizon...

and by the way, no disrespect meant, just a friendly debate

 

[TommyLeeBDW]

Don't know why people are still buying a android phone that cost $800 at launch and resale is worth only $200 two years later plus it is slower and have lower quality camera compare to an iphone at the same time of launch. I have many top of the line android phone and I always go back to iphone everytime due to lack of performance, resale and camera.

 

[Kenneth_79]

I'm the guy buying the former flagship phone when it is 2 years old. I'm worried how much will it cost for someone to replace the battery since most phones are sealed now.

 

[drmoriartiabroad86]

Wow.....Awesome banter on both sides..Thanks!!

 

[TEAMSWITCHER]

I didn't realize it was this bad... Many of my friends got the iPhone 8, even with the iPhone X on the way. Apple users seem to stick religiously to a two year upgrade cycle, because they can get decent value from their iPhone 6s trade-in.

 

[varase]

Has anyone considered that the operating systems are different and if you swapped the chips over to handle each other's OS that things may be different? Just a thought. I don't know enough to know if this could be a factor.

The reason there's stagnation is because of the business model of the product:

1. Google makes the OS, but doesn't profit that much from each phone.
2. OEMs make the phones, but outside of Samsung don't make much for each phone.
3. Qualcomm makes the chips, but they profit from each chip and each license.
4. Broadcom and others make chips, but only profit from each chip.

Each of the parties involved just make profit from pushing massive quantities of product out the door, and have no reason to make a superior long lasting future-proof device since their profits end pretty much when the device enters the supply chain.

Android users tend to be much less invested in pay services, or buying apps so Google's main benefit from Android is simply selling advertisements. If its good enough for the buyer to see the ad, it's good enough.

OEM handset makers would *love* to sell their users a handset every year or two, but most Android users are not flagship buyers and most fall into the realm of feature phone-priced handsets buyers and have little brand loyalty. Outside of custom brand-specific software (which Google discourages), there's not a whole lot to differentiate on brand from another.

The chipset makers just want to push out the maximum number of chips, and have little interest in a product which the user will be satisfied using for many years - and the chipset makers are only interested in incremental improvements which will encourage users to buy a new product next year, when new software will make the old product seem slow. A powerful future-proofed design would be a disaster because you would sell less chips over a user's lifetime.

Apple, on the other hand, actually has a reason to have the most powerful processors they can make. Having multiple year's handsets means they can cover different price points by offering the same handset whose manufacturing start-up cost have been depreciated away can now offer older handsets to cover a lower price points, thought that handset will have to compete with new lower cost handsets from Android OEMs.

 

[pcampbell804]

So, the A11 crushes it in benchmark scores. But not in real world use. I'm sorry, but as an engineer and former Apple software tester, if the benchmark score is the only aspect where significant improvement is found, then I would say something went wrong. And cherry picking a couple of limited examples does make a compelling argument for Apple's claims of 70% faster. Not when there are now dozens of side by side comparisons showing the iPhone 8 struggling to stay ahead of the iPhone 7, and not keeping up with a variety of competing phones. A short sample list is below:

"iPhone 8 vs Note 8" by Tech Trinkets: Note 8 win.
"iPhone 8 Plus vs iPhone 7 Plus" by MadMat: iPhone 8, barely.
"iPhone 8 Plus vs Sony Xperia XZ Premium" by Phone Battles: Sony win.
"iPhone 8 Plus vs OnePlus 5" by Phone Battles: OnePlus win.
"iPhone 8 Plus vs Nokia 8" by Tech Trinkets: Nokia win except for games.
"iPhone 8 Plus vs Galaxy Note 8" by SuperSaf TV: Note 8 win except for games.

If the A11 Bionic performed as suggested by benchmark scores, then none of these defeats or close-calls would exist.

The other aspect that bothers me as someone who has purchased every iPhone since gen 2, and currently own the iPhone 7 (128 GB, matte black, on AT&T), is that Apple has gone stale and fallen in the background for mobile media consumption. "Hey Tony, did you hear anything coming out of those earpods? Ya did? Ok, good enough". It's hard to think of Apple without including music and media in our thought process, yet now they are less than second rate, falling behind many others in audio quality. I was hoping, since I'm a professional recording engineer, that someone would come out with a dramatic breakthrough in audio that would rival standing in a studio control room, and go beyond analog headphone capabilities. You would expect Apple to be pioneering there, but who actually does come up with the breakthrough? HTC. The iPhone 8 may excel at very specific tasks, and benchmarks, but if real world use is insignificantly better than the previous model, and user experience is lower quality and less satisfying, does a synthetic score hold any weight?

 

[varase]

"iPhone 8 vs Note 8" by Tech Trinkets: Note 8 win.
"iPhone 8 Plus vs iPhone 7 Plus" by MadMat: iPhone 8, barely.
"iPhone 8 Plus vs Sony Xperia XZ Premium" by Phone Battles: Sony win.
"iPhone 8 Plus vs OnePlus 5" by Phone Battles: OnePlus win.
"iPhone 8 Plus vs Nokia 8" by Tech Trinkets: Nokia win except for games.
"iPhone 8 Plus vs Galaxy Note 8" by SuperSaf TV: Note 8 win except for games.

I took a look at your examples, and I'll bet what you're seeing is the effects of the new processor controller.

In an attempt to save battery, I'll bet the processor controller initially throws a lightweight engine (core) at a task and monitors its usage. If the usage exceeds a certain threshold, it throws another engine in and monit0rs those. In this way, processor engines are preserved and are ready to dispatch to another task should the need arise.

What this means is the old PhoneBuff trick of serially launching eight apps and round-robining them won't measure the true potential of the multiprocessing environment, as the dispatch scheme has radically changed to a "you gets resources if your process proves its needs" schema, rather than a "here's everything and the kitchen sink" schema.

Note that those serial launchings are in no way reflect any king of real-world usage.

That's not necessarily a bad way to dispatch engines -it just invalidates that simplistic kind of "real world benchmarking".

 

[dtzavell]

There is no way, 54% of 10.170 is 6.564. Do the math.